Journal of Environmental Management 144 (2014) 73e82

Contents lists available at ScienceDirect

Journal of Environmental Management journal homepage:

Is environmental management an economically sustainable business? Antje Gotschol a, Pietro De Giovanni b, *, Vincenzo Esposito Vinzi c a

Department of Information, Logistics and Innovation, VU Amsterdam University, de Boelelaan 1105, 1081 HV Amsterdam, the Netherlands Operations Management Department, ESSEC Business School, Avenue Bernard Hirsch, B.P. 105, 95021 Cergy Pontoise, Paris, France c Decision and Information Systems Department, ESSEC Business School, Avenue Bernard Hirsch, B.P. 105, 95021 Cergy Pontoise, Paris, France b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 September 2013 Received in revised form 19 January 2014 Accepted 1 May 2014 Available online

This paper investigates whether environmental management is an economically sustainable business. While firms invest in green production and green supply chain activities with the primary purpose of reducing their environmental impact, the reciprocal relationships with economic performance need to be clarified. Would firms and suppliers adjust their environmental strategies if the higher economic value that environmental management generates is reinvested in greening actions? We found out that environmental management positively influences economic performance as second order (long term) target, to be reached conditioned by higher environmental performance; in addition, firms can increase their performance if they reinvest the higher economic value gained through environmental management in green practices: While investing in environmental management programs is a short term strategy, economic rewards can be obtained only with some delays. Consequently, environmental management is an economically sustainable business only for patient firms. In the evaluation of these reciprocal relationships, we discovered that green supply chain initiatives are more effective and more economically sustainable than internal actions. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Green supply chain management Green production Environmental performance Economic performance Structural equation modeling

1. Introduction Environmental management refers to sustainable management approaches that aim for engaging in green processes and practices in order to reduce the environmental impact of the firm’s activities. This can be pursued on a firm-specific level within the framework of green production programs and also on a collaborative Supply Chain (SC) level that refers to environmental collaboration approaches (Rao, 2002; Theyel, 2000; Rao and Holt, 2005; De Giovanni, 2012; De Giovanni and Esposito Vinzi, 2012). Both types of programs have shown to substantially contribute to the environmental performance and therewith to the greenness of activities (Zhu and Sarkis, 2004; Zhu et al., 2005; Hervani et al., 2005). Most business and economic scientists agree that incorporating a green and environmentally sound strategy is crucial for businesses nowadays (e.g. Seuring and Müller, 2008; Zhu and Sarkis, 2004; Rao and Holt, 2005; Vachon and Klassen, 2006). While some studies are concluding with no significant results concerning the relationship between the economic and environmental performances (e.g. Rao, 2002; Jaggi and Freedman, 1982), numerous

* Corresponding author. Tel.: þ39 3398126474. E-mail address: [email protected] (P. De Giovanni). 0301-4797/© 2014 Elsevier Ltd. All rights reserved.

studies have been published reporting either positive (e.g. Alvarez Gillet al., 2001; De Giovanni, 2012) or negative (Bowen et al., 2001; De Giovanni and Zaccour, 2013, 2014) effects. This huge amount of research fails in the investigation of the non-recursive relationships between economic performance and environmental management. That is, how would firms and suppliers invest in environmental management if the higher economic performance generated by environmental management are reinvested in production and/or supply chain environmental initiatives? Yet, empirical research has failed to provide an answer to the issue regarding causality when reciprocal relationships exist. Although researchers such as Wagner et al. (2002) have attempted to account for the possibility of reverse causality in that concern, this topic has only been narrowly covered and asks for further assessment. Schaltegger and Synnestdevt (2002) highlight the difficulties that firms encounter in the engagement in environmental management programs. On one hand, the economic resources needed to invest in such programs can be substantially high, while the economic rewards are not at all guaranteed. Only profitable firms can afford to make sustainable investments in green activities to raise the environmental performance and face their environmental obligations (Stefan and Paul, 2008). On the other hand, firms do not


A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82

look at environmental practices as an opportunity to develop their business but rather as a cost-center (De Giovanni, 2012). This means that firms activate environmental programs more to conform with restrictions from regulations and legislations (e.g., Emission Trading) rather than extrapolating additional value through some atypical managerial practice (De Giovanni and Esposito Vinzi, 2014). If environmental management is economically viable, firms and suppliers will invest more in it. But, is there a loop between environmental management and economic performance? Answering to this question provides insights not only on the possibility to look at environmental management as a way to increase economic performance but also on the needs to continuously reinvest in green programs. To investigate these issues, the data of a survey on the engagement in and benefit from environmental management practice will be studied by applying Structural Equation Modeling (SEM). This paper is structured as follows. The second section includes relevant literature outlining the theory concerning environmental management and practices, as well as the different performance dimensions. Subsequently, the third section deals with the methodology. Besides more detailed information on the sample and the choice for structural equation modeling, this section includes an overview concerning data screening and assessment measures. This will be followed by the results section in which the outcome of the analysis is depicted. Lastly, the paper closes with a discussion on the results and a conclusions section.

2. Literature review and research hypotheses developments The awareness and conscious decisions to drive towards a larger degree of sustainability is increasing throughout businesses. Beyond governmental bodies, research showed that firms are forced to set their focus on environmental measures to prevent the negative impact of business activities (Linton et al., 2007; Bocken et al., 2011; Zhu and Sarkis, 2004; Vachon and Klassen, 2006). The demand for more sustainability and environmental friendliness has an increasing impact on businesses around the world, thus firms should consider environmental management as to be a value driver (De Giovanni and Zaccour, 2014). A different perspective captures sustainability with regard to opportunities for companies in terms of long-term risk reduction and economic performance (Shrivastava, 1995b). The latter can surely be obtained through adequate investments in production activities and supply chain collaboration (De Giovanni, 2012). In the pursuit of becoming more sustainable and reducing the environmental impact, firms can green their production process through several specific practices such as: the application of cleaner technologies as well as the setup of an environmental production strategy to reduce waste, emissions, and noise. As a consequence, production processes need to be adjusted in order to prevent and preserve the environment while facing the limitations linked to several barriers, for instance, usage of required materials for production (Sarkis, 2003), remanufacturing (Rao and Holt, 2005) or environmentally friendly packaging (Hervani et al., 2005). Cleaner production is a preventive strategy that pursues actions to eliminate or reduce waste and emission as well as to improve the energy flow while utilizing materials more efficiently (Fresner, 1998). A first approach to decrease the emissions is the minimization of water pollution and keeping the level of non-renewable energy resources on the lowest level that is feasible in accordance to the  rsson et al., 2009). Although green proproduction plan (Halldo duction programs are a relevant contribution to becoming more environmentally friendly, firms should also take into account some other operational criteria such as quality standards, accreditations

and certifications (Rao and Holt, 2005, Zhu and Sarkis, 2004), e.g. ISO14001 (GEMI, 1998) (Table 1). Significant support and incentives for developing green production practices can be given by suppliers belonging to the same SC. To align SC targets and the scope, firms can provide some incentives to suppliers to push investments in environmental programs and align all suppliers’ wishes to perform the environment. Supplier collaboration deals with the collaboration among the different supply chain parties (Vachon and Klassen, 2008), as a way to respond to the increasing environmental awareness over the supply chain, which will take the form of Green Supply Chain Management (GSCM) (Christopher and Ryals, 1999). Vachon and Klassen (2006) agree and state that organizations consider environmental strategies no longer on an individual single firm level, but on a collective SC level going beyond firm’s boundaries. In this perspective, De Giovanni and Zaccour (2014) report the type of incentive that a chain leader can provide to supply chain members to enhance their willingness to collaborate in environmental programs and contribute to GSCM. The vast majority of the literature in EM highlights the importance of cleaner production as an antecedence to GSCM (Theyel, 2000; De Giovanni, 2012; De Giovanni and Esposito Vinzi, 2012; Rao and Holt, 2005). In other words, a firm cannot be part of the GSCM, if it is not internally green. Green production represents a pre-requisite to be met by partners. While literature appears to be quite aligned on assessing the positive influence of internal environmental management on GSCM, we adopt a different perspective where the level of environmental collaboration undertaken with suppliers enhances firms’ green production investments. According to Fresner (1998) and De Giovanni and Zaccour (2014) environmental cooperation among SC partners is a fundamental condition to invest more in green production practices. To put light into this intuition, one can just ask a simple question: Why should a firm invest more in environmental management, if their suppliers do not commit to any form of green actions? The effectiveness of green investments could just be vanished when suppliers undertake non-green actions. Having environmental cooperation as a common and shared target over the supply chain builds trust and commitments among supply chain partners, which leads each firm to investment more in green production. Therefore, we can hypothesize that: H1 GSCM has a positive influence on green production. When it comes to how sustainability and greenness of a SC should be measured, different components should be taken into account. Being environmentally sustainable requires the company to be cautious about waste and pollution which are to be minimized. A sustainable business in this field incorporates the environmental and ecological approach not only on a long-term strategic level, but also on an operational level. For instance, De Giovanni and Zaccour (2013, 2014) highlight the benefits of implementing a GSCM in the form of closed-loop supply chain in which the return rate is a proxy of both, the environmental performance (low discard in environment) and economic performance (lower production cost due to the usage of returned components). Supply chain members expect the usage of transparent and welldefined standardized indicators to align objectives and targets with their partners (Ilinitch et al., 1999). The impact of environmental collaboration on environmental performance needs to be carefully investigated (Sarkis, 1999). While several (mainly empirical) studies illustrate that GSCM improves the environmental performance (e.g. Frosch, 1994; Geffen and Rothenberg, 2000; Green et al., 1996), other research did not support this

A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82


Table 1 Summary of the research hypotheses. Research hypothesis


Related research

H1. GSCM has a positive influence on Green Production.

Supplier collaboration on environmental programs pushes firms to undertake internal green programs and green their production processes. A firm is willing to invest more in green production if suppliers commit also some efforts to green the entire supply chain. Environmental cooperation becomes a driver to build trust over the chain: a firm trusts suppliers who are committed on environmental management and invests even more in green operations. Suppliers collaborate in the development of green programs with the primary objective to improve environmental performance. This is the core of a GSCM, in which firms collaborate to increase the capability of the entire chain to sustainably carry out their business. Firms undertake some green production programs to reduce the impact of their production process to the environment. This can be either imposed by suppliers as a fundamental criteria to be part of the chain or imposed by legislation that fixes some specific environmental targets. Suppliers collaborate in the development of green programs to improve economic performance only as second order target, while performing environment always is the primary objective. Higher economic performance can be achieved only eventually through improvements of environmental performance. Firms undertake some green production programs to improve economic performance only as second order target, while performing environment always is the primary objective. Higher economic performance can be achieved only eventually through improvements of environmental performance. The positive results obtained in the improvements of environmental performance translate into higher economic performance. For instance, a lower consumption of energy is directly linked to better environmental performance that leads, consequently, to higher economic performance through higher cost savings.

Fresner (1998), De Giovanni and Zaccour (2013, 2014)

H2. GSCM has a positive influence on environmental performance.

H3. Green production has a positive influence on Environmental Performance.

H4. GSCM has a positive and indirect influence on economic performance.

H5. Green Production has a positive and indirect influence on Economic Performance.

H6. Environmental performance has a positive influence on economic performance

H7. Economic performance has a positive influence on green production.

H8. Economic performance has a positive influence on GSCM.

When firms perform from an economic point of view due to the adoption of some green practices in production, firms reinvest the higher economic rewards in internal programs to generate a loop between green production and economic performance. When firms perform from an economic point of view due to the adoption of some collaborative practices with suppliers, firms reinvest the higher economic rewards in supplier collaboration to generate a loop between environmental collaboration and economic performance.

relationship (Wagner et al., 2002 and Levy, 1995). For instance, Savaskan et al. (2004) demonstrate that environmental collaboration with 30 service logistics provider never leads to higher environmental performance. Instead, collaboration with a retailer is always preferred when the maximization of environmental performance is in the menu. The foundations for this statement should be researched on the links between environmental collaboration and other strategies that a supplier undertakes. A recent report provided by the Port of Amsterdam ( highlights the importance of establishing collaborative initiatives to properly choose the logistics mode. Although collaboration boosts the business, higher CO2 emissions could be generated by a heavier

De Giovanni and Zaccour (2013, 2014), Ilinitch et al. (1999), Sarkis (1999), Frosch (1994); Geffen and Rothenberg (2000), Green et al. (1996), Wagner et al. (2002), Levy (1995), Savaskan et al. (2004), Wen-Hsien and Shih-Jieh (2009a,b), Zhu and Sarkis (2004), Zhu et al. (2005), Hervani et al. (2005).

De Giovanni (2011, 2012), De Giovanni and Esposito Vinzi (2012, 2014), De Giovanni and Zaccour (2013, 2014).

Zhu et al. (2005), Rao (2002), Rao and Holt (2005), Savaskan et al. (2004), Jacobs et al. (2010), Klassen and McLaughlin, (1996), Cordeiro and Sarkis (1997), De Giovanni (2011, 2012), De Giovanni and Esposito Vinzi (2012, 2014), De Giovanni and Zaccour (2013, 2014), Vachon and Klassen (2008), De Giovanni (2012), Rao and Holt (2005), Jacobs et al. (2010). Wen-Hsien and Shih-Jieh (2009a,b). Stefan and Paul (2008), Alvarez Gil et al. (2001), De Giovanni and Zaccour (2013, 2014), Hansmann and Kroeger (2001)

utilization of non-green transport mode (e.g. road transportation) with the result of damaging the environment. The adoption of green production and supplier collaboration practices has the main purpose of lowering the environmental impact of activities, while the improvement of economic performance should represent a second order target to be achieved only eventually. Although the positive impact of collaboration on environmental performance is supported by several studies (e.g., Vachon and Klassen, 2008; De Giovanni, 2012; De Giovanni and Esposito Vinzi, 2012, 2014), the consequences of environmentally friendly actions on the economic performance are not well understood and addressed. On one hand, Rao and Holt (2005) point


A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82

out that firms engaging in greening the business by introducing different initiatives will be remunerated only in the long run with reduced risk and cost, enhanced corporate image, as well as improved marketing advantages, that translate into higher sales and better economic performance (Jacobs et al., 2010). Minimizing waste will consequently also result in a better utilization of natural resources, improve the efficiency and lead to higher productivity which ultimately translates into decreasing operating cost. On the other hand, Zhu and Sarkis (2004) emphasize a stronger negative impact of EM on the economic performance due to higher investments and purchasing cost. Collaboration, joint environmental programs and common planning are time- and resource-intensive approaches that do not pay off instantaneously (Rao, 2002). Bowen et al. (2001) confirm that the positive impact of environmental performance on the economic performance cannot be attained within the short-term. Instead, firms engaging in greening the supply chain will provide the basis for long-term superior performance. Also, De Giovanni and Zaccour (2014) showed that the environmental performance is a short term target, while the economic performance (profits) is simply postponed. In order to contribute to the debate initiated in the literature, it is hypothesized that: H2. GSCM has a positive influence on environmental performance. H3. Green Production has a positive influence on environmental performance. H4. GSCM has a positive and indirect influence on economic performance. H5. Green production has a positive and indirect influence on economic performance. H6. Environmental performance has a positive influence on economic performance. Previous research has disregarded the economic sustainability of environmental management. That is, whether the higher economic value that green production and environmental collaboration with suppliers generate is re-invested in environmental initiatives. This presently represents an open question. The implementation of environmental initiatives passes through a sufficient amount of economic resources that both firms and suppliers should be able to invest, although economic improvements are not at all guaranteed. Profitable firms can afford to make sustainable investments in green activities to raise the environmental performance (Stefan and Paul, 2008). In fact, such investments, being positive with regard to environmental performance, may even negatively affect the economic and financial performance (Alvarez Gil et al., 2001). Hence, an economic evaluation of environmental initiatives discourages investments due to their impact on the economic performance (De Giovanni and Zaccour, 2013). Longterm benefits such as competitive advantage or emerging new market opportunities and other value adding approaches (Hansmann and Kroeger, 2001) seem to be ignored by a number of businesses. This research seeks to investigate whether firms should depart from the traditional and short term view of looking at environmental management as cost center to evaluate the benefits that it exerts in the long term through reinvesting the higher economic value generated. If a loop exists between environmental management and economic performance, investing in green production and environmental collaboration is economically sustainable. Therefore, it is hypothesized that: H7. Economic Performance has a positive influence on green production. H8. Economic Performance has a positive influence on GSCM.

Fig. 1. Conceptual model.

Research hypotheses are displayed and summarized in Fig. 1.

3. Methodology 3.1. Scale developments 3.1.1. Green production The literature refers to green production as a successful concept to achieve greenness and sustainability at the stage of manufacturing (e.g. Rao and Holt, 2005; Lewis, 2000). This practice helps not only to improve operationally, but more importantly, it reduces the environmental impact within the production phases (Lewis, 2000). Nunes and Bennett (2010) depict a variety of production related activities including their environmental relevance. When aiming towards a greener production previous studies highlight the importance of product design and process technology, since these have a significant impact on waste and energy consumption (Sarkis, 1995; Shrivastava, 1995a). Similar achievements are possible through technological innovation (Angell and Klassen, 1999) and innovative equipments (Shrivastava, 1995a). Cleaner technologies allow for a more efficient usage of natural resources through less harmful components and lower air, water, and soil pollution (OECD, 1995). Therefore, we used environmentallyfriendly raw materials (GrProd1), substitution of environmentally questionable materials (GrProd2), consideration of environmental criteria (GrProd3), optimization of processes to reduce air emissions, water use, solid waste, and/or noise (GrProd4), and use of cleaner technology process to make savings (GrProd5) to assess the dimension Green Production.

3.1.2. GSCM Suppliers undertake programs and actions to reduce the energy used and emissions over the supply chain (Parry et al., 2007; Bloemhof-Ruwaard et al., 1995; Hervani et al., 2005; Zhu et al., 2005; Theyel, 2000). GSCM is a new ‘archetype’ for businesses that still focuses on common goals (e.g., profit and market share) but simultaneously includes environmental impacts and risks (van Hoek, 1999). This concept intends to approach a SC-wide collaboration to entirely green the SC (Rao and Holt, 2005; Zhu et al., 2005; Zhu and Sarkis, 2004; Vachon and Klassen, 2008). GSCM aims towards a joint planning, common sustainable practices and collaborative decision-making, to minimize the environmental impact (Vachon and Klassen, 2006; Zhu et al., 2005; van Hoek, 1999) and align actions of all SC participants. Rao (2002) describes the support of the larger company in the SC encouraging and guiding their suppliers through the implementation of green practices. Suppliers may also provide incentives to firms to align their processes to the GSCM targets (De Giovanni and Zaccour, 2013) or even impose penalties when green investments are not sufficiently high. More

A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82


environmental performance used over the research are: reduction of air emissions (EnvP1), reduction of solid/liquid wastes (EnvP2), and reduction of the amount of energy used (EnvP3), decrease of consumption for hazardous/harmful/toxic materials (EnvP4), and decrease of frequency for environmental accidents (EnvP5).

Table 2 Sample composition. Categories of sample composition

Distribution (%)

SC manager Purchasing manager Production manager Material manager Other titles

50 25 13.33 8.33 3.33

Sector of firms Food Glass Cement and ceramics Metal Paper and pulp Industrial equipment Housing Electronic products Mechanics Textiles Others Annual gross sales (values of 2006) Range in the sample Average in the sample

16.25 13.33 12.5 11.6 10 9.16 7.1 7.1 6.66 5 1.3 in V 15 millionse3.4 billions 98 millions

general, partnering and mentoring practices between SC partners contribute towards successful environmental improvements. In order to capture all these aspects, we wish to investigate the construct GSCM through the following indicators: guiding suppliers to establish their own environmental programs (GSCM1), choosing of suppliers by environmental criteria (GSCM2), achievement of environmental goals collectively (GSCM3), development of a mutual understanding of responsibilities regarding environmental performance (GSCM4), work together to reduce the environmental impact of the SC (GSCM5), conduct of joint planning to anticipate and resolve environment-related problems (GSCM6). 3.1.3. Environmental performance Environmental performance requires a peculiar measurement approach that accounts for non-financial performance (Ilinitch et al., 1999). For instance, Ilinitch et al. (1999) use a categorical system of four components, namely internal systems, external stakeholder relations, external impact, and internal compliance. While these categories cover a broad range of factors, previous research has mainly focused on more precise components such as air emissions, water, soil pollution, or toxic releases (e.g. Rao, 2002). The literature also contains some ad hoc indicators. GEMI (1998) developed an eco-productivity index for a pharmaceutical company to measure the degree of its environmental performance and that includes the use of raw materials, the water and energy consumption, as well as packaging. De Giovanni and Zaccour (2013) introduced an indicator of environmental performance that takes into account the component's lifespan and the supply chain members' green activity programs. This heterogeneity in measuring environmental performance has pushed literature to measure environmental performance through a latent variable, to capture several aspects of the same phenomenon, that is, the reduction of the environmental impact. Therefore, indicators of

3.1.4. Economic performance Firms are interested in the economic performance which is about maximizing profits, minimizing costs and increasing the market shares. Earlier research considered scales such sales growth, net income growth and profitability (ROI) for their measurement purposes, which is still applicable to later research that referred to variables such as profit, sales, growth margins (Rao, 2002) as well as classically cost savings or market share (Kim, 2006). Rao and Holt (2005) e besides the classical economic indicators e consider variables of new market opportunities and products. By contrast, however, Rao and Holt (2005) link the classical indicators with other performance indicators that are also subject to the environmental performance, e.g. improved productivity. The link between a firm’s environmental investments and indicators that relate to the economic performance is a crucial step towards the assessment of interrelationships between the different performance indicators. Maxwell et al. (2006) focused on four main indicators to analyze: Cost benefits resulting from eco-efficiencies (and thereby constructing the link to environmental performance), competitive advantage, market share and productivity. In order to capture the firm's capability of producing economic value, the dimension of economic performance is investigated through the following scale: profit (EcP1), market share (EcP2), and cost savings (EcP3). 3.2. Sample description To answer our research questions, the hypotheses will be tested through primary data specifically collected for the purposes of this study. Managers (1400) have been contacted and asked to fill in a questionnaire. All selected organizations are Italian and are registered in the Analisi Informatizzata delle Aziende (AIDA, Italian company information and business intelligence) database. This database e hosted by the Bureau Van Dijk e provides information on companies from different sectors and with different sizes on a global, regional as well as domestic level. Thereby, fields concerning financials, corporate structures and other business intelligence are covered. Within the framework of this questionnaire, interviewees were asked to answer questions concerning the engagement on environmental management. In order to assess the content validity as well as the suitability of the chosen variables, experts (faculty members in the field of SC and logistics management) were requested to check the content of the questionnaire. The content validity of the survey as well as its relevance to the study has been assessed by four SC managers. The appropriateness of the sample for this research has been ensured by firstly inquiring whether the respective firm is part of a SC and secondly whether it also engages in green production and

Table 3 p-Values on the non-respondent bias tests.

Early vs late respondents Supply chain manager vs others

Early vs late respondents Supply chain manager vs others












0.492 0.932

0.202 0.624

0.379 0.792

0.930 0.335

0.285 0.681

0.494 0.398

0.069 0.264

0.682 0.139

0.077 0.471

0.415 0.727

0.053 0.129








0.409 0.869

0.054 0.116

0.093 0.306

0.470 0.786

0.202 0.562

0.531 0.587

0.496 0.269

0.883** 0.903** 0.084 0.111 0.090 0.002 0.093 0.072 0.057 0.068 0.061 0.645** 0.633** 0.640** 0.788** 0.005 0.084 0.594** 0.582** 0.584** 0.703** 0.725** 0.007 0.017 0.554** 0.611** 0.581** 0.440** 0.529** 0.512** 0.039 0.110 0.520** 0.493** 0.522** Correlation significance: ***Significant at a  0.01; **Significant at a  0.05; *Significant at a  0.1 (two-tailed test).

0.572** 0.448** 0.482** 0.543** 0.091 0.064 0.509** 0.492** 0.471** 0.688** 0.617** 0.458** 0.504** 0.548** 0.088 0.018 0.511** 0.487** 0.508** 0.192** 0.295** 0.094 0.068 0.095 0.033 0.077 0.019 0.123 0.052 0.083 0.195** 0.100 0.084 0.044 0.124 0.082 0.107 0.027 0.008 0.008 0.040 0.023 0.103 0.240** 0.579** 0.605** 0.453** 0.379** 0.441** 0.451** 0.103 0.074 0.393** 0.358** 0.371** 0.341** 0.013 0.080 0.444** 0.462** 0.429** 0.390** 0.470** 0.516** 0.012 0.057 0.563** 0.566** 0.531** 0.712** 0.392** 0.088 0.012 0.479** 0.479** 0.397** 0.504** 0.537** 0.634** 0.031 0.052 0.568** 0.591** 0.555** 0.805** 0.731** 0.428** 0.045 0.076 0.482** 0.491** 0.436** 0.504** 0.570** 0.610** 0.042 0.062 0.577** 0.585** 0.549** 0.128* 0.083 0.023 0.013 0.168** 0.074 0.008 0.027 0.032 0.121 0.094 0.078 0.124 0.185** 0.011 0.006 0.047


Table 4 Correlation matrix.

Before testing the research hypotheses H1eH8 through structural equation modeling (SEM), a confirmatory factor analysis (CFA) was run (see Table 5) to identify a statistically suitable final model. To obtain a good fit of the model, it was necessary to drop some of the items included in the measurement scales as they resulted in a standardized coefficient of less than 0.70. Items that contributed to standardized residuals with values greater than 3.00 were also deleted. Although these measurement variables were not used in the further analysis, they are still reported and highlighted in the Appendix 1. Four constructs (Green production, GSCM, environmental performance and economic performance) were obtained within the CFA. All indicator variables that are kept in the model fulfill necessary requirements and are therefore considered meaningful. The standardized value or factor loadings (l) are above 0.7 and are statistically significant with a p-value 0.05. Each linkage between variables is assessed by applying the critical ratio. By considering a confidence interval of 95% as the respective significance level, each critical ratio exceeding the 1.96 magnitude is referred to as being significant. In the present CFA analysis, the lvalues met the requirements and therefore adding to the convergent validity. In addition to that, for each indicator variable the R-squared value is reported. It provides information on the


3.3. Model assessment

0.050 0.802** 0.806** 0.709** 0.403** 0.014 0.094 0.483** 0.463** 0.430** 0.485** 0.529** 0.583** 0.049 0.046 0.570** 0.583** 0.549**

environmental collaboration with suppliers. The details on the survey’s questions are supplied in Appendix 1. The sample firms received the questionnaire including an additional cover letter with information concerning the purpose of this research. After three weeks, 110 usable responses were obtained resulting in a first response rate of approx. 7.85%. After a second letter (leading to further 43 responses) and finally directly contacting remaining firms by phone (resulting in another 87 responses), the response rate eventuating in a value of 17.1%, meaning 240 firms answered. Response rates in that scope are considered sufficient and are common in this field of research (e.g. Zhu et al., 2005; Melnyk et al., 2003; Vachon and Klassen, 2008). The obtained sample is regarded as suitable. As outlined in Table 2, it comprises a wide variety of top-level managers, different industries, as well as different sizes (defined in annual gross sales). The industry distribution closely reflects the composition of the initial choice of firms whom the questionnaire was initially sent. This allows for a higher degree of generalizability. To check the representativeness of the sample with respect to the population, we have conducted a one-sample chi-square test on the sector of firms. This test allows for examining whether the proportions of the sample firms in each sector are similar to or significantly different from the population proportions in each sector. We used the weighted cases method to determine the extent to which the sample (n ¼ 240) generalizes to the population (N ¼ 1.400). Our results show a chi-square value of 14.740 (df ¼ 10) with a p-value ¼ 0.142, thus the proportions in the sample are not significantly different from the proportions in the population and the sample is then representative of the population. In order to detect a possible non-response bias, it has been tested for differences in answers between early respondents and late respondents. A conducted one-way ANOVA test revealed no significant differences between these two types of respondents. The same holds for the different manager groups where no significant difference in answers could be proven. Table 3 reports the p-values of tests run on each indicator. In Table 4, all variables that are relevant for the construct development are reported including their respective correlations.


A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82



A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82


Table 5 Confirmatory factor analysis results. Factor

Indicator Variable

Estimated Loading

Standardized value (l) loading

Standard Error (S.E.)

Critical Ratio (C.R.)


Green Production

GrProd1 GrProd2 GrProd3

1.048 1.033 1.000

0.905 0.889 0.806

0.064 0.065 e

16.274 15.953 e

0.819 0.790 0.649



1.216 1.235 1.025 1.000

0.836 0.824 0.721 0.691

0.107 0.110 0.102 e

11.373 11.246 10.038 e

0.699 0.679 0.520 0.478

Environmental performance

EnvP1 EnvP2 EnvP3

0.864 0.907 1.000

0.799 0.862 0.916

0.053 0.048 e

16.261 18.715 e

0.638 0.743 0.839

Economic performance

EcP1 EcP2

1.017 1.000

0.940 0.939

0.044 e

23.025 e

0.883 0.883

variance that is explained by the respective predictor, thereby by default also indicating the percentage of error variance (1  R2) the observed variable accounts for concerning the variance of the variable itself. The standard error (S.E.) is reported per respective estimated value. In order to assess the model fit and allow for an appropriate elaboration, different fit indexes are regarded (see Table 6 for a comprehensive summary). The final evaluation of all measurements led to the conclusion that the measurement model is of acceptable fit, meaning that the observed variables do reliably reflect the theoretical construct. Within this model the chi-square has a value of 59.09, with a p-value of 0.131. The results imply that the model is acceptable, as the assessed deviation of the estimated variance-covariance matrix to the sample variancecovariance matrix is found to be statistically non-significant. Additional indexes provide further information concerning the model fit. Summarizing, according to all outlined indexes and model fit assessments, the presented model is considered to have a satisfactory model fit, hence providing an appropriate basis for further analyses approaches. Besides indexes, reliability (convergent validity), average variance extracted, and discriminant validity have been checked (Fornell and Larcker, 1981). TheCronbach’s a is reported as means for internal consistency and validity. All Cronbach's a were above the established threshold of 0.7 to be considered satisfactory in terms of reliability (Nunnally, 1978), ranging from 0.850 to 0.938. Second, the average variance extracted (AVE) is considered as a discriminant validity criterion. The AVE assessment, proposed by Fornell and Larcker (1981), measures the shared variance of constructs. The AVE should exceed a value of 0.5 to be satisfactory. This condition has been met by all constructs with AVE values of 0.594 to 0.883. Moreover, Table 7 includes a factor correlation matrix with the square root of the AVE on the diagonal. By comparing the AVE value of each construct with the squared correlation of the respective construct with any other, it results that no critical discriminant validity issues

emerged. Third, the construct reliability e as a measure for internal consistency e is expected to be above 0.8. In order to examine the latter requirement, the composite reliabilities of all constructs were considered. Those ranged from 0.853 to 0.938 and thereby met the requirement. 3.4. Structural equation modeling Using IBM SPSS Amos 21, we tested the research hypotheses HeH8 through SEM. The fit statistics of the structural model indicate a good model fit, meaning that the model reproduces the data of the sample in an appropriate manner. The chi-square is at a level of 59.09, with 48 degrees of freedom, leading to a chi-square to df ratio (c2/df) of 1.23. The chi-square statistics is non-significant with a p-value above 0.05 (exact value: 0.131). The GFI (0.961) and AGFI (0.936) both meet the requirements of a good fit. In addition to that, the TLI (0.993), CFI (0.995) and RMSEA (0.031) comply with the thresholds of the respective indicator, leading to the conclusion that the structural model has a good fit. Furthermore, the results of the conducted analysis reveal an adequate measurement portion of the model. From Table 8, Hypotheses H1 and H2 are corroborated by the results of this study. Both paths are statistically significant with a pvalue 0.05 c2/df < 3 is good; c2/df < 5 is sometimes permissible GFI >0.95 AGFI >0.80 TLI >0.95 CFI >0.95 is good; CFI >0.90 is traditional; CFI >0.80 is sometimes permissible RMSEA 0.10 is not acceptable




A. Gotschol et al. / Journal of Environmental Management 144 (2014) 73e82

Table 7 Scale properties and correlations. Construct

No. of Indicators

Composite Reliabilitya


Factor correlations GrProd


Green production (GrProd) 3 0.901 0.753 0.868 GSCM 4 0.853 0.594 0.655 0.771 Environmental performance (EnvP.) 3 0.895 0.740 0.724 0.721 Economic performance (EcP.) 2 0.938 0.883 0.699 0.650 P P P a Composite reliability ¼ ½ð standardized loadings2 Þ=½ standardized loadings2 þ εj . P P P b Avg: Variance Extracted ¼ ½ ðstandardized loadingsÞ2 =½ ðstandardized loadingsÞ2 þ εj  where εj is the indicator measurement errors.

of green production on environmental performance with a standardized positive factor loading of 0.339. With a ¼ 0.007, the result is statistically significant, leading to H3 as being supported by this analysis. Companies that engage in green production contribute towards an improved environmental performance of the business. Results with regards to the impact that the environmental performance has on the economic performance are found to be statistically significant with a strong factor loading of 0.537, substantiating hypothesis H6. Environmental performance is positively associated with the economic performance e thereby providing support for the predominant opinion amongst researchers concerning the relationship of environmental and economic performance. The analysis results concerning H4 and H5 which focus on the indirect effect that green production and GSCM actions may have on the economic performance of a company uphold the expectation of a positive influence from both practices. Both hypotheses are confirmed. Green production showed a standardized indirect effect of 0.223, while the results imply a stronger impact of GSCM on economic performance with an indirect loading of 0.341. With reference to the causal loops of this model (H7 and H8), the results signify a positive link back to green production and GSCM actions. The reciprocal impact of the economic performance on green production activities is statistically significant with a p-value

Is environmental management an economically sustainable business?

This paper investigates whether environmental management is an economically sustainable business. While firms invest in green production and green sup...
373KB Sizes 5 Downloads 5 Views