at SciVerse ScienceDirect

Journal of Cleaner Production 47 (2013) 129e140

Contents lists available

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

Environmental management and operational performancein automotive companies in Brazil: the role of human resourcemanagement and lean manufacturing

Charbel José Chiappetta Jabbour a,*, Ana Beatriz Lopes de Sousa Jabbour a,Kannan Govindan b, Adriano Alves Teixeira c, Wesley Ricardo de Souza Freitas d

aUNESP, Univ Estadual Paulista (The Sao Paulo State University), Avenida Engenheiro Edmundo Carrijo Coube, Bauru, São Paulo State,CEP 17033360, BrazilbUniversity of Southern Denmark, Department of Business and Economics, Odense 5230, Denmarkc Federal University of Mato Grosso do Sul, Paranaiba, BR497, KM12, MS 79500-000, BrazildUSP, University of Sao Paulo, Avenida Bandeirantes, 3900, Ribeirao Preto, Sao Paulo State, CEP 14040905, Brazil

a r t i c l e i n f o

Article history:Received 14 March 2012Received in revised form21 June 2012Accepted 5 July 2012Available online 25 July 2012

Keywords:Environmental managementLean manufacturingHuman resource managementOperational performanceAutomotive sectorBrazil

* Corresponding author.E-mail addresses: cjabbour@terra.com.br,

(C.J.C. Jabbour).

0959-6526/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.jclepro.2012.07.010

a b s t r a c t

The main objective of this study is to verify the influence of Environmental Management (EM) onOperational Performance (OP) in Brazilian automotive companies, analyzing whether LeanManufacturing (LM) and Human Resources (HR) interfere in the greening of these companies. Therefore,a conceptual framework listing these concepts was proposed, and three research hypotheses werepresented. A questionnaire was elaborated based on this theoretical background and sent to respondentsoccupying the highest positions in the production/operations areas of Brazilian automotive companies.The data, collected from 75 companies, were analyzed using structural equation modeling. The mainresults are as follows: (a) the model tested revealed an adequate goodness of fit, showing that overall, therelations proposed between EM and OP and between HR, LM and EM tend to be statistically valid; (b) EMtends to influence OP in a positive and statistically weak manner; (c) LM has a greater influence on EMwhen compared to the influence HR has over EM; (d) HR has a positive relationship over EM, but thestatistical significance of this relationship is less than that of the other evaluated relationships. Theoriginality of this paper lies in its gathering the concepts of EM, LM, HR and OP in a single study, as theygenerally tend not to be treated jointly. This paper also provided valid empirical evidence for a little-studied context: the Brazilian automotive sector.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The intensification of environmental concerns has been leadingcompanies to adopt environmental management practices at anincreasing rate (Boiral, 2006; González-Benito, 2006). One of thearguments favoring the adoption of these environmentalmanagement practices is that they can benefit firms, giving rise tothe so-called “green and competitive” (Porter and Van Der Linde,1995; Hunt and Auster, 1990; Berry and Rondinelli, 1998; Molina-Azorin et al., 2009). Among those benefits that can be ascertainedfrom environmental management is the improvement in firms’operational performance, such as a reduction in production costs(Porter and Van Der Linde, 1995). However, specialized literature

prof.charbel@gmail.com

All rights reserved.

affirms that environmental management can create synergy withmanagement practices from other areas in a firm (Wagner, 2007).

Two management areas have gained prominence as targets ofeffective environmental management (Wilkinson et al., 2001). Thefirst is operations/manufacturing management, which, because itprocesses resources, has significant environmental effects. Thesecond area is human resources, which may influence the perfor-mance of new organizational objectives, such as those related toenvironmental performance.

The ability of the operations/manufacturing area to supportenvironmental management tends to be greater when thecompany adopts Lean Manufacturing practices (González-Benitoand González-Benito, 2008). This type of relationship has becomeknown as the “Lean and Green” hypothesis and has been analyzedby several authors (Simpson and Power, 2005; Rothenberg et al.,2001; King and Lenox, 2001; Yang et al., 2011; Maxwell et al.,1998). These authors argue that, in general, waste reduction in

Table 1Variables related to environmental management.

Environmentalmanagement (EM)variables/practices

Measures/definition Source

Clear policy ofvalorizing environmentalmanagement (EM1)

Clear policy of valorizationof environmental managementthrough a precise declarationfrom business directors aboutthe main environmentalaspects and impacts generated.

Boiral (2006)

Environmental training forall employees (EM2)

Environmental training for allemployees aimed at promotingenvironmental policy andpermitting employee awarenessof their activities’ environmentalimpacts.

Daily andHuang (2001)

3Rs (Reduction, Reuse andRecycling applied towater, electric energyand paper) (EM3)

3Rs, comprising Reduction,Reuse and Recycling appliedto water, electric energy, paperand other natural inputs,increasing business productivity.

Marcus andFremeth (2009)

Development of productswith smallerenvironmentalimpacts (EM4)

Development of products withsmaller environmental impacts.

Sarkis (2001)

Development of productionprocesses with smaller

Development of productionprocesses with smaller

Sarkis (2001)

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140130

manufacturing contributes to environmental management(Simpson and Power, 2005) through greater efficiency in the use ofproduction resources (Rothenberg et al., 2001) and the adoption ofcleaning practices and improved organization of the productiveenvironment (King and Lenox, 2001), which can generatecompetitive advantages (Yang et al., 2011).

On the other hand, the support of human resource managementpractices is also considered fundamental for adopting environ-mental management practices (Jackson et al., 2011; Govindarajuluand Daily, 2004; Sarkis et al., 2010). These researchers affirm thathuman resource management must align its practices (such asrecruiting, selection, performance evaluation, and training) withenvironmental management objectives. This process is calledGreen Human Resource Management (Renwick et al., 2008), whichfollows the hypothesis that a more intense alignment betweenhuman resources and environmental issues leads more firms toadopt environmental management practices (Bohdanowicz et al.,2011; Jabbour et al., 2010).

However, there are no studies which integrate EnvironmentalManagement (EM), Operational Performance (OP), LeanManufacturing (LM) and Human Resources (HR). There are fewstudies that partially investigate these relationships. For example,Jabbour et al. (2012) analyze the relationship between environ-mental management and operational performance; May andFlannery (1995) investigate the relationship between environ-mental management and human resources; Rothenberg et al.(2001) analyze the relationships between lean manufacturing andenvironmental management. There is thus an opportunity forresearch that fully analyzes this relationship. Ideally, this relation-ship should first be verified in the automotive industrial sector,which is considered by some researchers (Womack et al., 2004) tobe a pioneering industry for management practices and tendencies.Brazil was chosen as the country of analysis due to the growinginterest of its researchers in environmental management as well asthe high relevance of the automotive sector in the country’s GDP.

Therefore, this study’s main objective is to verify the influence ofenvironmental management on the operational performance ofBrazilian automotive companies, analyzing whether leanmanufacturing and human resource management play a role in thegreening of these companies. Based on this objective, this papertests a conceptual framework based on structural equationmodeling. In the face of the other statistical techniques available,structural equation modeling is advantageous because (a) itpermits researchers to test more complex conceptual frameworks,guaranteeing a more robust and holistic statistical analysis (Ismailet al., 2012), and (b) it permits the simultaneous analysis of therelationships between a broad range of variables (Hair et al., 2011).

The following sections of this paper introduce the study’sconceptual framework with its respective research hypotheses(Section 2). This study also details the methodological proceduresused for collecting and analyzing data (Section 3), presents theresults and discusses them in light of the literature (Section 4) and,in the conclusion, discusses the main implications of this study anddescribes a proposal for future studies (Section 5).

environmental impacts(EM5)

environmental impacts.

Supplier selection based onenvironmental criteria(EM6)

Vendor selection based onenvironmental criteria.

Jabbour andJabbour (2009)

ISO 14001 or otherEnvironmentalManagementSystem (EM7)

Environmental managementsystems (ISO 14001and/or others).

ABNT NBRISO 14001(2004)

Voluntary promotion ofinformation onenvironmentalperformance (EM8)

Voluntary promotion ofinformation on environmentalperformance.

Boiral (2006)

2. Research hypotheses and conceptual framework

According to Haden et al. (2009), environmental managementconcerns the complete incorporation of environmental objectivesand strategies to the broader objectives and strategies pursued bythe organization. Jabbour (2010) complements this definition,suggesting that environmental management be based ona systemic approach incorporating environmentally consciousstrategy at every level of the organization.

Several factors can lead a company to adopt environmentalmanagement practices (Berry and Rondinelli, 1998). According toGonzález-Benito (2006), stakeholder pressure is the main factordriving organizations toward more advanced environmentalmanagement. More advanced environmental management canalso improve a company’s financial performance (Molina-Azorinet al., 2009) and increase the company’s manufacturingcompetitiveness, promoting cost reductions, quality improve-ments and the generation of new products and processes (Yanget al., 2010).

In addition, especially with the advances of the population’senvironmental awareness, companies that invest in environmentalmanagement may increase in worth through green marketinginitiatives (Woolverton and Dimitri, 2010). Another means ofincreasing worth occurs when organizations announce theiradoption of ISO 14001 environmental management systems, whichtends to generate an increase in share value traded on stockexchanges (Jacobs et al., 2010).

There is thus an emerging consensus in the literature (Darnallet al., 2008; Iraldo et al., 2009; Crowe and Brennan, 2007; Vachonand Klassen, 2008; Yang et al., 2010; González-Benito, 2005;Sroufe, 2003) that there are positive results correlating the adop-tion of environmental management practices with the organiza-tions’ performance, gauged through various indicators, especially atenvironmentally proactive organizations.

It is believed that the adoption of these environmentalmanagement practices (Table 1) may generate advantages inseveral measures of operational performance in organizations

Table 3Human resource practices.

Human resources (HR)variables/practices

Measures/definition Source

Recruiting andselection (HRM1)

Recruiting consists of attractingnew people to company, andselection consists of choosingthe right people for a job.

Dessler (2003)

Training (HRM2) A planned organizational actionthat permits acquiring technicaland behavioral skills whilecontributing to the developmentof cognitive strategies that canmake the individual more apt toperform current or future functions.

Borges-Andrade(2002)

PerformanceEvaluation (HRM3)

Process that aims to determine anemployee’s work results; one of itsmain functions is to offer a reasonfor compensating his results andefforts.

Türk andRoolaht (2007),Stoner andFreeman (1999),Robbins andDecenzo (2004)

Rewards (HRM4) The term refers to all monetarypayments and all goods ormerchandise used to rewardemployees.

Daft (1999),Hipolito (2002)

Benefits (HRM5) These are the benefits andconveniences shared by theorganization and by employeesthat are not part of the directsalary.

Oliveira andLeone (2008),Bateman andSnell (1998)

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140 131

(Table 2), here including costs, quality, flexibility, delivery, newproduct development and time-to-market for new products.

Thus, the first hypothesis of this study is: H1 e The adoption ofenvironmental management practices has a positive correlationwith the organizational performance of companies in Brazil’sautomotive sector.

However, environmental management should be com-plemented with the management of human and behavioralelements supporting environmental management practices,which have been gaining strength (Perron et al., 2006) andrequire support from human resource management (Jacksonet al., 2011), through the HR practices (Table 3) in pursuit ofenvironmental management objectives. This study assumes thatmore efficient and effective resource management practices leadthe human resources field to better understand the organization’sobjectives and goals and to be better able to contribute toachieving these goals (Collins and Clark, 2003). In this sense,Huselid et al. (1997) affirm that more effective human resourcepractices are associated with higher organizational performancebecause their human resource departments are better equippedto contribute to the achievement organizational goals. Osmanet al. (2011) affirm that human resource practices are positivelyrelated to the performance of Malaysian firms as well. Therefore,organizations’ human resources e as well as practices for theirproper management when efficient and with efficacy e areessential drivers of a sustained competitive advantage (Jamrogand Overholt, 2004; Voorde et al., 2010).

In the specialized literature, this process of support from HR toEM objectives is called Green Human Resource Management(GHRM) (Renwick et al., 2008). GRHM concerns the alignment ofseveral practices in human resource management (recruiting,selection, training, performance evaluation, rewards, etc.) witha company’s environmental management objectives (Renwicket al., 2008; Muller-Carmem et al., 2010).

Table 2Variables related to operational performance.

Operationalperformance(OP) variables

Measures/definition Source

Cost (OP1) Seeks the lowest pricecompared to competitors,the lowest total productioncost, or the highest productioncapacity.

Hayes and Wheelwright(1984), González-Benito(2005), González-Benito(2006)

Time-to-Market(OP2)

Refers to the time needed toplace a product in a market, thatis, from conception to availabilityat the final point of sale.

González-Benito(2005), González-Benito(2006)

New Products(OP3)

Entry of products into a specificmarket aiming to attract newconsumers and/or retainingcurrent ones. Related to productswith new characteristics andfunctionalities.

González-Benito(2005), González-Benito(2006)

Quality (OP4) Zero-defect manufacturing ormanufacturing of durableproducts.

Hayes and Wheelwright(1984), González-Benito(2005), González-Benito(2006)

Flexibility(OP5)

Quick changes in product design,quick introduction of newproducts, quick changes inproduction volume, broad varietyof products, or quick changes inproduct mix.

Hayes and Wheelwright(1984), González-Benito(2005), González-Benito(2006)

Delivery(OP6)

Quick delivery or reliability intimely deliveries.

Hayes and Wheelwright(1984), González-Benito(2005), González-Benito(2006)

Some practical results confirm the importance of HR for EM: (a)Sarkis et al. (2010) conducted a survey with 157 large companies inSpain’s automotive sector. They concluded that environmentaltraining is a mediating variable for the success of environmentalmanagement practices in analyzed companies; (b) Jabbour et al.(2010) observed in a survey with 94 Brazilian companies thatmore evolved environmental management leads to more supportfrom human resource practices.

We thus present H2 e The adoption of human resourcemanagement practices has a positive correlation with the envi-ronmental management of companies in Brazil’s automotive sector.

According to King and Lenox (2001), the logic of organizationand cleaning in lean production practices has the benefits of wastereduction and a lower risk of accidents. Maxwell et al. (1998) foundimplementations of lean production to be dedicated to a philos-ophy of waste reduction that could be easily understood to achievethe objectives of environmental protection. Vais et al. (2006)suggest that to become lean and environmentally friendly, theorganization should focus on energy consumption and materialresidue, which are the inputs and outputs of a transformationsystem. In this context, it can be stated that the adoption of leanproduction practices improves the organization’s environmentalperformance.

Yang et al. (2011) state that it is important for manufacturingcompanies to implement lean production practices with environ-mental management as a means of obtaining eco-advantagesthrough improvements in environmental performance. Accordingto Dües et al. (2013), companies can use lean practices as a catalystfor greening the supply chains because “lean” and “green” haveoverlapping practices and elements.

Some studies reported how lean manufacturing practices(Table 4) can positively influence actions geared toward corporateenvironmental management. Maxwell et al. (1998), Rothenberget al. (2001) and Simpson and Power (2005) stressed the impor-tance of involving employees, whether to intervene in the processto avoid failures (that cause rework and unnecessary use ofresources) or to commit to and propose improvements related to

Table 4Variables related to lean production practices.

Lean manufacturing(LM) Variables/Practices

Measures/definition Source

Multifunctionalinvolvement inthe process(LM1)

Development of employeeskills and incentive forautonomy to avoid failuresthroughout the process.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Continuousimprovement(LM2)

Seeks incremental continuousimprovement in quality, costs,delivery and the project.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

5S (LM3) A form of visual managementfor reducing disorder andinefficiency in the productiveand administrative environments.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Total productivemaintenance(LM4)

Aims to improve machinereliability and capacity throughperiodic maintenance regimes.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Kanban (LM5) Card system for creating apulled flow.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Just-in-Time(LM6)

Seeks a continuous productionflow.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Lot reduction/stockreduction (LM7)

Formation of small productionlots to reduce stock in processand to increase variety.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Improvementcircles/kaizencircles (LM8)

Promote systematic discussionsbetween operators and managersfor better incremental continuousimprovement.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Vendordevelopment/collaboration(LM9)

Activities geared towarddeveloping relationshipswith the vendor to obtaintheir collaboration.

Biazzo and Panizzolo(2000), Shah andWard (2003), Bhasinand Burcher (2006),Pettersen (2009).

Operational Performance (OP)

Lean Manufacturing (LM)

Environmental Management Practices (EM)

Human Resources (HR)

H1

H2

H3

Fig. 1. Research framework.

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140132

the improved use and conservation of inputs. In this same sense,quality circles are another form of promoting this involvement, asthey provide employees with training and workshops intended tomotivate them to participate in projects for environmental effi-ciency and to engage in responsible consumption (Vais et al.,2006).

Vais et al. (2006) also cite the 5S and total productive mainte-nance as lean production practices that aid environmentalmanagement. The 5S provide guidelines to the organization andcleaning to avoid the incorrect disposal of waste and incorrect useof inputs. Total productive maintenance aims at the periodic reviewof equipment based on simple adjustments (cleaning, lubrication,calibration, etc.) to increase the useful life of equipment and itsefficiency (Donaire, 1999).

According to Pojasek (2008), lean production practices adhere toseveral ISO 14001 standards. Examples of such practices areseeking the root cause of a problem and thus applying correctiveactions, creating conditions for preventing failures (jidoka/pokayoke) and thus elaborating emergency action procedures, andproviding continuous improvement based on critical analysis by

top management. King and Lenox (2001) and Rothenberg et al.(2001) verified that high levels of pollution prevention occur atplants that use lean production practices, among other reasons, dueto stock reductions. Another prominent factor according toSimpson and Power (2005) and Corbett and Klassen (2006) is theimportance of supplier collaboration in the environmentalimprovement process, as suppliers are responsible for providinginputs that directly affect the environmental efficiency of the finalproduct.

This study thus proposes H3 e The adoption of leanmanufacturing practices has a positive correlation with the envi-ronmental management of companies in Brazil’s automotive sector.

From a review of the extant literature, a conceptual frameworkis proposed and shown in Fig. 1.

This theoretical framework was empirically tested following themethodological procedures below.

3. Methodology

3.1. Methodological framework

Based on the existing gap in research combining environmentalmanagement, operational performance, human resources and leanmanufacturing applied to the Brazilian context, it was decided toconduct a quantitative study. This approachwas chosen because forall the individual concepts analyzed, quantitative scales alreadyexist in specialized literature. It thus was possible to conducta survey. Fig. 2 shows the flow of procedures and methodologicalchoices for this survey.

3.2. Industrial sector studied

The Brazilian automotive sector, specifically the auto partssector, is the target of this study. Brazil’s automotive sector began inthe 1950s and has since evolved into 26 car manufacturers with 53factories supplied by more than 5000 auto part companies, with aninstalled production capacity of 4.3 million vehicles and 109thousand farm machines per year, positioning Brazil as one of thesix biggest producers of vehicles in Brazil (Anfavea, 2011).

Based on data from 2010 it is possible to affirm that the sectoremploys approximately 1.5 million people, earns more than US$107.6 billion annually (including auto parts), has a total productiontotaling 5.2% of Brazil’s gross domestic product (GDP), and canreach 22.5% of GDP if all indirect effects are considered (Anfavea,2011).

This extensive growth in the sector should not be attributedonly to the car manufacturers because it was made possible by the

Research Framework

Empirical Test

Methodological Procedures

Respondents: Operations

Directors/Managers

75 valid questionnaires

Use of Structural Equation Modeling

Results/Discussions

Hypotheses Test

Model Test

Final Considerations

Gap in literature and Brazilian context

Objective of the study

Hypotheses H1, H2 and H3

Survey Study

Brazilian Automotive Sector

Elaboration and Test of the Questionnaire

Fig. 2. Flow of procedures and methodological choices for this survey.

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140 133

installation of the auto part industry, which together with themanufacturers was responsible for several innovations, such asflex-fuel engines and other technological adaptations to the Bra-zilian market.

3.3. Questionnaire building

A data collection instrument was planned for collecting databased on a structured questionnaire about the concepts previouslyreviewed in Section 2 that were elaborated according to therecommendations contained in Synodinos (2003).

The questionnaire contains information on the characterizationof respondent companies and four blocks of assertions: one for the“Environmental Management” construct, one for “LeanManufacturing”, one for “Human Resources” and the last for“Operational Performance”. Altogether, the questionnaire presentseight assertions about environmental management (one for eachenvironmental practice), nine about the lean manufacturingconstruct, five about human resource management and six aboutoperational performance (one for each measure of operationalperformance). The first version of the questionnaire was submittedfor content validation through an analysis of five researchers in thearea, as well as the adjustment to conceptual presuppositions. In itsfinal version, the questionnairewas hosted in a virtual environmentspecifically elaborated for this research.

A 5-point Likert scale was adopted, where 1 represents “totallydisagree” and 5 represents “totally agree”.

3.4. Data collection

The research data were collected between October 2010 andMarch 2011. First, e-mail addresses and telephone information for654 automotive sector companies (auto part segment) located inBrazil were collected at the National Automotive Vehicle Compo-nent Industry Union. E-mails were sent to respondents occupyingthe highest positions in production/operations areas at Brazilianautomotive companies. The e-mails contained a brief explanationabout the study and an invitation for the operations/manufacturingmanager to participate. The choice of the operations/manufacturing manager was made because the operations/manufacturing area generates most of the environmental impactsand is responsible for several operational performance measures. Itis also the area responsible for LM practices. In addition, becausethe operations/manufacturing manager is a line manager, heshould be familiar with the HR practices for managing operations/manufacturing area employees.

The e-mail contained a link to direct the target respondentdirectly to the questionnaire hosted in the study’s virtual envi-ronment. Phone calls were also made to increase the return of validquestionnaires, and an attempt was made to contact the employeesresponsible for the company’s production area.

A total of 72 questionnaires was collected through the researchsite, and 4 questionnaires were collected from alternative means, asrequested by the respective respondents. In all, 76 questionnaireswere obtained, 1 of which was discarded due to being incomplete,leading to a total response rate of 11.11% (75 valid questionnaires),a number considered adequate compared to the percentages sug-gested by Synodinos (2003) and Large and Thomsen (2011).Murillo-Luna et al. (2011) state that response rates greater than 6%can already be considered adequate for attempts at extrapolatingresults, especially in studies that apply structural equationmodeling. As will be seen in the next section, goodness of fit (GoF),a general adjustment indicator for the statistical model, achievedgood scores for this study, which also indicates that the sample wasadequate (see Section 4). Each filled-out questionnaire automati-cally fed a data spreadsheet for subsequent statistical processing.

3.5. Analysis of results

The conceptual framework (presented in Section 2) guided thedata analysis process, which involved the use of statistical proce-dures with the support of data spreadsheets from the StatisticalPackage for Social Sciences (Version 19.0) and SmartPLS 2.0. Section4 presents the statistical procedures associated with each of theresults obtained in detail and shows a consolidation table vali-dating or rejecting the study’s hypotheses.

The statistical tests involved the following boundaries forapplication:

� Adjustment of the sample for each individual factor using theKMO (Kaiser-Meyer-Olkin) test. The KMO test verifies thecorrelation value between the variables, and if the value issmall, the KMO test near zero the sample can be inadequate. Onthe other hand, a value close to 1 can be considered adequate(Hair et al., 2005);

� Using Principal Components Analysis to group variables intofactors (Hair et al., 2005);

� Calculating Cronbach’s Alpha for each factor. Cronbach’s Alphais used to measure construct reliability. Reliability is under-stood as the measure of internal consistency of responsesbetween respondents for a single construct (Kline, 2005);

� Bartlett’s Test of Sphericity. Bartlett’s test evaluates thehypothesis that the correlation matrix is the identity matrix,

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140134

where the determining factor equals one (Pestana and Gageiro,2003). This test is used to analyze the correlation matrix asa whole;

� Main diagonal of the anti-image matrix, which should presentvalues greater than 0.6. (Hair et al., 2005);

� Variable communalities, which explain the adherence ofa given variable to the diverse factors of a factorial analysis(Hair et al., 2005);

� The eigenvalues for each factor, fromwhich factors with valuesequal to or greater than 1.0 were extracted. A factor’s eigen-value indicates howmuch data cloud variance is absorbed by it(Aranha and Zambaldi, 2008);

� R2 values near 0.75, 0.50 and 0.25 are considered substantial,moderate and weak, respectively (Hair et al., 2011);

� The t test value near 1.65, 1.96 and 2.58 are considered withsignificance levels of 10%, 5% and 1%, respectively (Hair et al.,2011);

� GoF (goodness of fit statistics), which measures the overallstatistical fitness of the model tested, can have values of GoF-small ¼ 0.1; GoF-medium ¼ 0.25; GoF-large ¼ 0.36 (Wetzelset al., 2009).

The application of these statistical measures will be shownbelow as part of the presentation of the empirical test of theproposed conceptual model.

4. Results and discussions

The reduction of data for all variables from the EnvironmentalManagement (EM) construct, the Lean Manufacturing (LM)construct, the Human Resources (HR) construct and the Opera-tional Performance (OP) construct was performed using PrincipalComponent Analysis (PCA) through the varimaxmethod (Appendix1e4).

In relation to the Environmental Management Construct (EM),only one factor was formed, explaining an accumulated variance ofapproximately 74.38%, with an eigenvalue of 5.95 and proper valuesin the main diagonal of the anti-image matrix (0.848; 0.821; 0.925;0.863; 0.852; 0.951; 0.930; 0.908). The KMO test, which assessessample fitness, was 0.882, an adequate level, as was the value ob-tained with the Bartlett Test of Sphericity (636.937, andp value< 0.1) and Cronbach’s Alpha (0.949). All of the EM Constructvariables presented satisfactory values (Appendix 1A).

After refining the Environmental Management Construct (EM)reported above, the EM1 (environmental policy) variablewas foundto obtain the highest average among the environmental manage-ment practices (Appendix 1B). The Pearson coefficient of correla-tion test was also run (Appendix 1C), revealing that all EM1eEM2variables have significant correlations, underscoring the relationbetween EM1 (environmental policy) and EM2 (environmentaltraining).

Therefore, environmental management at the analyzedcompanies tends to constitute the totality of practices consideredherein, confirming the indications by González-Benito (2006)concerning the implementation of environmental managementthrough a set of practices. The “environmental policy” practicestood out with the highest average, as did the “environmentaltraining” variable, which had the highest coefficient of correlationwith the environmental management construct; the importance ofthis correlation has been emphasized by several authors(Govindarajulu and Daily, 2004; Daily and Huang, 2001; Sarkiset al., 2010).

Concerning the Human Resources (HR) construct, only onefactor was formed, explaining an approximate accumulated vari-ance of 68.12%, with an eigenvalue of 2.72 and values adjusted in

the main diagonal of the anti-image matrix (0.71; 0.61; 0.68; 0.63).The KMO test, which verifies sample fitness, produced a value of0.662, which is considered to be an adequate level, as are the valuesobtained from the Bartlett Test of Sphericity (141.41, andp value < 0.1) and Cronbach’s Alpha (0.84). The Human ResourcesConstruct (HR) comprised the variables HRM1, HRM2, HRM3 andHRM4. Variable HRM5 was excluded due to low communality(0.38) (Appendix 2A).

After refining the Human Resources Construct (HR) reportedabove, the variable HRM2 e training e was found to obtain thehighest average among human resource practices (Appendix 2B).The Pearson coefficient of correlation test was also run, revealingthat all HRM1eHRM4 variables have significant correlations,underscoring the relation between HRM1 (recruiting and selection)and HRM2 (training) (Appendix 2C).

As a consequence, human resource management in the autoparts sector tends not to adopt a homogenous standard ofbenefits for employees, as HRM5 was not statistically valid. Thisreveals sector specificity, which is consistent with the contin-gency approach of human resource management suggested byJackson and Schuler (1995). On the other hand, most humanresource management practices found in the literature reviewwere verified in practice, especially HRM2 (training practice),which makes the worker more apt to perform daily work activi-ties at an industrial establishment as suggested (Borges-Andrade,2002).

For the Lean Manufacturing (LM) construct, only one factor wasused. This factor explained an accumulated variance of approxi-mately 64.27%, with an eigenvalue of 5.78 and proper values in themain diagonal of the anti-image matrix (0.917; 0.904; 0.927; 0.903;0.867; 0.841; 0.943; 0.908). The KMO test, which assesses samplefitness, was 0.900, which is considered adequate, as are the valuesobtained with the Bartlett Test of Sphericity (460.202, andp value< 0.1) and Cronbach’s Alpha (0.927). All of the LM Constructvariables presented satisfactory values (Appendix 3A).

After refining the Lean Manufacturing (LM) construct reportedabove, the variable LM2 e Systematic Search for ContinuousImprovementewas found to obtain the highest average among LMpractices (Appendix 3B). The Pearson coefficient of correlation testwas also performed (Appendix 3C), revealing that all LM1eLM9variables have significant correlations, underscoring the relationbetween LM5 (Kanban) and LM6 (Just-in-Time).

Therefore, the Lean Manufacturing construct is perceived tohave all variables validated. Among all the practices, the “system-atic search for continuous improvement” obtained the highestimplementation average and was also the most important variablein the structural model test for the Lean Manufacturing construct.In terms of correlation, interdependence was verified among allLean Manufacturing variables, underscoring the relationshipbetween LM5 (Kanban) and LM6 (Just-in-Time). This correlationcan be explained by the importance of Kanban systems for imple-menting Just-in-Time (Ohno, 1988).

The Operational Performance (OP) construct comprised vari-ables OP1, OP2, OP5 and OP6. Variables OP3 and OP4were excludedfrom the analysis because they present communalities of 0.38 and0.43, respectively (Appendix 4A).

After refining the Operational Performance Construct (LM) re-ported above, the variable OP6 (capacity for meeting deadlinesestablished by clients) was found to obtain the highest averageamong operational performance practices (Appendix 4B). ThePearson coefficient of correlation test was also performed(Appendix 4C), revealing that all OP1, OP2, OP5 and OP6 variableshave significant correlations; the correlation between OP5 (flexi-bility for adapting to clients) and OP6 (capacity to meet clientdeadlines) is particularly significant.

Table 5Reliability and validity values for the structural model.

Constructs Averagevarianceextracted(AVE)

Compoundedreliability

R2 Cronbach’salpha

Communality

EM 0.743662 0.958608 0.39569 0.950406 0.743662HR 0.677598 0.893229 0.000 0.84124 0.677598LM 0.641557 0.941442 0.000 0.930237 0.641557OP 0.661935 0.886552 0.114243 0.831282 0.661935

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140 135

As a consequence, the configuration of the Operational Perfor-mance construct was only partially validated. This finding indicatesthat there is no clear perception that sector company performanceis measured in terms of launching new products or in terms ofdifferentiation in quality. This result can be explained by the factthat auto part products tend to follow launch specifications and thequality established by the car manufacturers. Furthermore, qualityhas become a qualifying factor and not awinner of supply contracts.

Next, structural equation modeling e Partial Least Squares(SEM-PLS)ewas used. Structural equationmodeling through PLS isconsidered a second-generation multivariate analysis. It is espe-cially useful whenworking with complex theory (relating conceptssuch as EM, OP, HR and LM) or in initial stages of development. Astructural model was created containing the constructs obtainedfrom Principal Component Analysis, as explained above (Fig. 3). Theanalyses were conducted using SmartPLS 2.03 (Sosik et al., 2009).

HR and LM were observed to positively influence EM with an R2

of 0.396, that is, with a moderate to weak intensity, according toHair et al. (2011). In this relationship, LM is most prominent and isthe most important construct explaining EM behavior. OP is posi-tively but weakly influenced by EM, as shown in the R2 value of0.114.

Good quality indicators for the proposed model were achievedin terms of Average Variance Extracted (convergent validity),compounded reliability, Cronbach’s Alpha and communalities, forall constructs. To assess satisfactory reliability (which identifies theprecision with which the construct measures exactly what isintended to be measured) and validity (which tests the relationshipof one variable with another variable from a same construct), thecompounded reliability value should be greater than 0.7, whereasthe convergent validity value should be greater than 0.5. Constructreliability was evaluated using compounded reliability. The

Fig. 3. Structur

convergent validity was analyzed using the Average VarianceExtracted. Table 5 shows that all compounded reliability values aregreater than 0.7 and that all Average Variance Extracted values aregreater than 0.5 (Foltz, 2008). The Cronbach’s Alpha coefficientsand the communalities are also considered adequate.

One means of guaranteeing discriminant validity is to assesswhether the variables do in fact have higher loads in their factors oforigin. This analysis obtained adequate results (Table 6).

Aimed at testing model robustness, a bootstrap of 1000 sub-samples was used to estimate the statistical significance of rela-tionships between proposed variables and constructs (Fig. 4).According to the Methodology section, when the value of the t testis close to 1.65, 1.96 and 2.58, the significance levels will be,respectively, 10%, 5% and 1% (Hair et al., 2011).

Therefore, the relationship between LM and EM is positive andsignificant at the 1% level. The same is valid for the relationshipbetween EM and OP. Therefore, environmental management tendsto influence operational performance in a positive but weakmanner, although with statistical significance. Finally, it is worthunderscoring the relationship between HR and EM, which provedpositive but with a significance level of only 10%.

al model.

Table 6Crossed loads for evaluating discriminant validity.

EM HR LM OP

EM1 0.90 0.32 0.55 0.32EM2 0.93 0.39 0.54 0.35EM3 0.87 0.36 0.53 0.26EM4 0.82 0.48 0.51 0.24EM5 0.86 0.44 0.59 0.32EM6 0.81 0.42 0.48 0.21EM7 0.87 0.36 0.54 0.33EM8 0.83 0.33 0.45 0.29HRM1 0.26 0.81 0.47 0.45HRM2 0.38 0.85 0.48 0.50HRM3 0.42 0.89 0.42 0.39HRM4 0.37 0.73 0.35 0.42LM1 0.51 0.43 0.80 0.44LM2 0.46 0.38 0.84 0.41LM3 0.52 0.35 0.82 0.23LM4 0.48 0.38 0.81 0.30LM5 0.35 0.30 0.72 0.24LM6 0.42 0.36 0.79 0.28LM7 0.43 0.47 0.81 0.38LM8 0.56 0.51 0.83 0.43LM9 0.58 0.51 0.78 0.39OP1 0.32 0.41 0.38 0.82OP2 0.19 0.35 0.27 0.75OP5 0.30 0.55 0.40 0.85OP6 0.26 0.40 0.35 0.84

Table 7Significance of model relationship coefficients.

Relationship Load T test Significance level

EM1 ) EM 0.90 37.84 *EM2 ) EM 0.93 61.57 *EM3 ) EM 0.87 20.52 *EM4 ) EM 0.82 15.31 *EM5 ) EM 0.86 25.77 *EM6 ) EM 0.81 18.74 *EM7 ) EM 0.87 33.69 *EM8 ) EM 0.83 17.92 *HRM1 ) HR 0.81 10.41 *HRM2 ) HR 0.85 16.01 *HRM3 ) HR 0.89 37.15 *HRM4 ) HR 0.73 8.80 *LM1 ) LM 0.80 16.14 *LM2 ) LM 0.84 21.74 *LM3 ) LM 0.82 19.91 *LM4 ) LM 0.81 22.36 *LM5 ) LM 0.72 8.98 *LM6 ) LM 0.79 13.73 *LM7 ) LM 0.81 17.80 *LM8 ) LM 0.83 19.38 *LM9 ) LM 0.78 15.24 *OP1 ) OP 0.82 11.17 *OP2 ) OP 0.75 5.10 *OP5 ) OP 0.85 10.83 *OP6 ) OP 0.84 9.87 *EM / OP 0.34 3.75 *HR / EM 0.18 1.84 ***LM / EM 0.52 6.21 *

*p value <0.01; **p value <0.05; ***p value <0.1 (Scale based on Hair et al., 2011).

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140136

All of the other model relationships are statistically valid at thesignificance level (p value) lower than or equal to 0.01, as perTable 7.

Finally, the GoF (Goodness of Fit Statistics) for the statisticalmodel should be determined. According toWetzels et al. (2009), forstudies in which the average R2 is close to 0.25, GoF should havea minimum value of 0.36 (GoF-large). In this study, the average R2

Fig. 4. Structural model with bootst

was found to be 0.255, and the average GoF was 0.443, indicatinggood fit. This finding indicates the proposed model overall hasa fitting statistical adjustment.

Therefore, the main hypothesis for this study that EM and OPcan be considered to be valid, with the observed relationship

rapping of 1000 sub-samplings.

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140 137

having a significance level of 1%, indicates that EM does indeed tendto influence OP among sample companies. However, this relation-ship tends to be weak. The validation is corroborated by the classicmantra in the literature, “green and competitive”, which emergedin the 1990s (Hunt and Auster, 1990; Porter and Van Der Linde,1995; Berry and Rondinelli, 1998) and has recently been taken upagain (Marcus and Fremeth, 2009; Jacobs et al., 2010; Molina-Azorin et al., 2009). However, this relationship, being weak,should be analyzed further by researchers.

The hypothesis relating LM and EM was also supported; therelationshipwas found to be positive at the highest level of statisticalsignificance. This relationship was found to be the most significantone identified in the model, revealing a positive and valid relation-ship between lean manufacturing and environmental managementpractices and thus confirming that companies can create synergiesbetween “lean” and “green” actions (Dües et al., 2013).

Finally, the relationship between HR and EM can also be consid-ered positive but only at a significance level of 10%. As suggested byHair et al. (2011), this hypothesis can be considered valid but witha statistical reliability that is less than for the model’s other hypoth-eses. This result, despite requiring more in-depth and qualitativeanalyses to gain knowledge, may be explained by the phenomenonwhereby companies generally forget the human side of environ-mental management (Jackson et al., 2011; Daily and Huang, 2001).

5. Conclusions

The objective of this study was to verify whether environmentalmanagement influences operational performance at Brazilianautomotive companies. It also verified whether environmentalmanagement is influenced by human resource management andlean manufacturing.

The combination of these themes in a single conceptualframework and empirically testing it in the context of Braziliancompanies is the primary contribution of this study. It is possible tofind some studies dedicated to investigating only part of thisframework of relationships, such as human resources and envi-ronmental management (Daily et al., 2012), but the opportunityremains to study more complete models such as the one presentedhere.

The main results of this study show that, in general, theconceptual model is statistically valid for those companiesanalyzed, as it results in a GoF of 0.423 (the cutoff line was 0.36,according to Wetzels et al., 2009). The empirical analysis alsorevealed the following:

� EM tends to influence OP in a positive and statistically valid (pvalue <0.01) but with a weak explanatory power. This findingindicates that relationship must be strengthened within thecompanies studied to generate synergy between environ-mental management and performance, creating, winewinconditions.

� LM tends to influence EM in a positive and statistically valid (pvalue <0.01) but weak-to-moderate manner. LM was found tobe the variable with the most explanatory power over EM.

� HR tends to influence EM in a positive manner, but this rela-tionship can only be accepted with a less rigorous statisticalcondition (p value < 0.1), which can be maintained with someexceptions. This finding indicates that HR does not have thesame significance power that LM has over EM.

These results have implications for scholars and business ownersalike. For scholars, in light of the Brazilian context, the literature’semphases on “green and competitive” (Porter and Van Der Linde,1995) and “lean and green” (Florida, 1996) are confirmed, but the

green human resource management approach was not found to besignificant (Jackson et al., 2011) for the analyzed companies.

For business managers, the main implications are as follows: (a)there is a need to systematically understand the relationshipbetween diverse approaches and managerial practices, and (b) thereis a need to pay more attention to the human side of environmentalmanagement, which can improve operating performance. These twomanagerial recommendations can contribute to those organizationsthat have been seeking more sustainable social repositioning.

These results may be useful for professionals dedicated toteaching environmental management, human resources or leanmanufacturing. They can also be useful for subjects concerning“doing business in Brazil” as well as those relating to internationalbusiness.

The main limitations of this research are sample size, which,despite all the effort made on data collection, only included 75participating companies, and the restriction of analyzing a singleindustrial sector. Another limitation concerns the existence ofoverlaps between HR, EM and LM variables, as discussed by Düeset al. (2013). Finally, it is believed that future studies are neededto better understand the reasons for the poor integration betweenhuman resource management practices and environmentalmanagement practices at analyzed companies.

Acknowledgements

This research is partially supported by the FAPESP e The SaoPaulo Research Foundation (Research Process # 2011/23454-1). Theauthor Kannan Govindan is supported by a grant from Forsknings-og Innovationsstyrelsen for “The International Network pro-gramme e Sustainable supply chain management: A step towardEnvironmental and Social Initiatives” (2211916).

Appendix 1

Appendix 1A. Result of the Principal Component Analysis for EM

Variables Load Communalities

EM1

0.900 0.811 EM2 0.936 0.875 EM3 0.874 0.764 EM4 0.818 0.669 EM5 0.850 0.723 EM6 0.808 0.653 EM7 0.875 0.766 EM8 0.831 0.690

In: All variables.

Appendix 1B. Average and standard deviation for the EM Construct

Variables Average Standard deviation

EM1

3.24 1.71 EM2 3.08 1.68 EM3 3.02 1.48 EM4 2.96 1.47 EM5 3.17 1.39 EM6 2.57 1.41 EM7 3.04 1.81 EM8 2.65 1.58

Appendix 1C. Pearson correlation for the EM Construct variables

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140138

EM1 EM2 EM3 EM4 EM5 EM6 EM7 EM8

EM1

1 EM2 0.939* 1 EM3 0.717* 0.781* 1 EM4 0.624* 0.660* 0.746* 1 EM5 0.687* 0.689* 0.783* 0.836* 1 EM6 0.655* 0.678* 0.653* 0.645* 0.659* 1 EM7 0.854* 0.899* 0.676* 0.608* 0.605* 0.628* 1 EM8 0.700* 0.773* 0.666* 0.548* 0.623* 0.684* 0.737* 1

*p value <0.05.

Appendix 2

Appendix 2A. Result of the Principal Component Analysis for HR

Variables Load Communalities

HRM1

0.85 0.72 HRM2 0.86 0.74 HRM3 0.87 0.76 HRM4 0.70 0.5

Out: Variable HRM5 (low communalitie/load).

Appendix 2B. Average and standard deviation for the HR Construct

Variables Average Standard deviation

HRM1

3.48 1.08 HRM2 3.52 1.10 HRM3 3.20 1.30 HRM4 2.34 1.12

Appendix 2C. Pearson correlation for the HR Construct variables

HRM1 HRM2 HRM3 HRM4

HRM1

1 HRM2 0.71* 1 HRM3 0.58* 0.72* 1 HRM4 0.49* 0.35* 0.55* 1

*p value <0.05.

Appendix 3

Appendix 3A. Result of the Principal Component Analysis for LM

Variables Load Communalities

LM1

0.79 0.63 LM2 0.84 0.71 LM3 0.81 0.65 LM4 0.81 0.66 LM5 0.74 0.54 LM6 0.81 0.65 LM7 0.82 0.67 LM8 0.82 0.68 LM9 0.75 0.57

Appendix 3B. Average and standard deviation for the LM Constructvariables

Variables Average Standard deviation

LM1

3.69 1.12 LM2 3.86 1.05 LM3 3.78 1.18

(continued )

Variables

Average Standard deviation

LM4

3.20 1.27 LM5 2.90 1.41 LM6 3.04 1.42 LM7 3.52 1.01 LM8 3.20 1.37 LM9 3.17 1.18

Appendix 3C. Pearson correlation for the LM Construct variables

LM1 LM2 LM3 LM4 LM5 LM6 LM7 LM8 LM9

LM1

1 LM2 0.737* 1 LM3 0.637* 0.708* 1 LM4 0.627* 0.653* 0.671* 1 LM5 0.499* 0.479* 0.518* 0.535* 1 LM6 0.512* 0.568* 0.546* 0.657* 0.771* 1 LM7 0.600* 0.643* 0.528* 0.534* 0.615* 0.701* 1 LM8 0.598* 0.688* 0.638* 0.579* 0.530* 0.581* 0.677* 1 LM9 0.524* 0.567* 0.590* 0.592 0.419* 0.497* 0.627* 0.647 1

*p value <0.05.

Appendix 4

Appendix 4A. Result of the Principal Component Analysis for OP

Variables Load Communalities

OP1

0.79 0.64 OP2 0.78 0.62 OP5 0.84 0.71 OP6 0.84 0.70

Out: Variables OP3 and OP4 (low communalitie/load).

Appendix 4B. Average and standard deviation for the OP Constructvariables

Variables Average Standard deviation

OP1

4.25 0.89 OP2 4.11 0.89 OP5 4.34 0.74 OP6 4.36 0.78

Appendix 4C. Pearson correlation for the OP Construct variables

OP1 OP2 OP5 OP6

OP1

1 OP2 0.58* 1 OP5 0.51* 0.52* 1 OP6 0.53* 0.50* 0.70* 1

*p value <0.05.

References

ABNT NBR ISO 14001/2004, 2004. Sistema de Gestão Ambiental. Especificação ediretrizes para uso. Associação Brasileira de Normas Técnicas, ABNT, Rio deJaneiro. versão.

Anfavea, 2011. Associação Nacional dos Fabricantes de Veículos Automotores.Available at: www.anfavea.org.br. Accessed June 2011.

Aranha, F., Zambaldi, F., 2008. Análise fatorial em administração. Cengage Learning,São Paulo.

Bateman, T.S., Snell, S.A., 1998. Administração: construindo vantagens competitivas.Trad. Celso A. Rimoli. Atlas, São Paulo.

Berry, M.A., Rondinelli, D.A., 1998. Proactive environmental management: a newindustrial revolution. The Academy of Management Executive 12 (2), 38e50.

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140 139

Bhasin, S., Burcher, P., 2006. Lean viewed as a philosophy. Journal of ManufacturingTechnology Management 17 (1), 56e72.

Biazzo, S., Panizzolo, R., 2000. The assessment of work organization in leanproduction: the relevance of theworker’s perspective. IntegratedManufacturingSystems 11 (1), 6e15.

Bohdanowicz, P., Zientara, P., Novotna, E., 2011. International hotel chains andenvironmental protection: an analysis of Hilton’s we care! Programme (Europe,2006e2008). Journal of Sustainable Tourism 19, 797e816.

Boiral, O., 2006. Global warming: should companies adopt a proactive strategy?Long Range Planning 39, 315e330.

Borges-Andrade, J.E., 2002. Desenvolvimento de medidas em avaliação de treina-mento. Estudos de Psicologia 7, 31e43. número especial.

Collins, C.J., Clark, K.D., 2003. Strategic human resource practices, top managementteam social networks, and firm performance: the role of human resourcepractices in creating organizational advantage. Academy of Management Jour-nal 46, 740e751.

Corbett, C.J., Klassen, R.D., 2006. Extending the horizons: environmental excellenceas key to improving operations. Manufacturing & Service Operations Manage-ment 8 (1), 5e22.

Crowe, D., Brennan, L., 2007. Environmental considerations within manufacturingstrategy: an international study. Business Strategyand theEnvironment16, 266e289.

Daft, R.L., 1999. Administração. Trad. Fernando Gastaldo Morales, fourth ed. LTC, Riode Janeiro.

Daily, B., Huang, S., 2001. Achieving sustainability through attention to humanresource factors in environmental management. International Journal of Oper-ations & Production Management 21 (12), 1539e1552.

Daily, B.F., Bishop, J.W., Massoud, J.A., 2012. The role of training and empowermentin environmental performance: a study of the Mexican. International Journal ofOperations & Production Management 32 (5), 631e647.

Darnall, N., Henriques, I., Sadorsky, P., 2008. Do environmental managementsystems improve business performance in an international setting? Journal ofInternational Management 14 (4), 364e376.

Dessler, G., 2003. Administração de recursos humanos. Tradução Cecília LeãoOderich, second ed. Pearson Prentice Hall, São Paulo.

Donaire, D., 1999. Gestão ambiental na empresa. Atlas, São Paulo.Dües, C.M., Tan, K.H., Lim,M., 2013. Green as thenew lean: how touse leanpractices as

a catalyst togreeningyour supplychain. Journalof CleanerProduction40, 93e100.Florida, R., 1996. Lean and green: the move to environmentally conscious

manufacturing. California Management Review 39 (1), 80e105.Foltz, C.B., 2008. Why users (fail to) read computer usage policies. Industrial

Management & Data Systems 108 (6), 701e712.González-Benito, J., 2005. A study of the effect of manufacturing proactivity on

business performance. International Journal of Operations & ProductionManagement 25 (3), 222e241.

González-Benito, J., 2006. Environmental proactivity and business performance: anempirical analysis. Omega: The International Journal of Management Science33, 1e15.

González-Benito, J., González-Benito, O., 2008. Operations management practiceslinked to the adoption of ISO 14001: an empirical analysis of Spanish manu-facturers. International Journal of Production Economics 113, 60e73.

Govindarajulu, N., Daily, B.F., 2004. Motivating employees for environmentalimprovement. Industrial Management & Data Systems 104 (4), 364e372.

Haden, S.S.P., Oyler, J.D., Humphreys, J.H., 2009. Historical, practical, and theoreticalperspectives on green management: an exploratory analysis. ManagementDecision 47 (7), 1041e1055.

Hair Jr., J.F., Babin, B., Money, A.H., Samouel, P., 2005. Fundamentos de métodos depesquisa em administração. Bookman, Porto Alegre.

Hair, J.F., Ringle, C.M., Sarstedt, M., 2011. PLS-SEM: indeed a silver bullet. Journal ofMarketing Theory and Practice 19 (2), 139e151.

Hayes, R.H., Wheelwright, S.C., 1984. Restoring Our Competitive Edge e Competingthrough Manufacturing. John Wiley & Sons, USA.

Hipolito, J.A.M., 2002. Sistema de recompensas: uma abordagem atual. In:Fleury, M.T.L. (Ed.), As pessoas na organização. Gente, São Paulo.

Hunt, C.B., Auster, E.R., 1990. Proactive environmental management: avoiding thetoxic trap. MIT Sloan Management Review 31 (2), 7e18.

Huselid, M.A., Jackson, S.E., Schuler, R.S., 1997. Technical and strategic humanresource management effectiveness as determinants of firm performance.Academy of Management Journal 40 (1), 171e188.

Iraldo, F., Testa, F., Frey, M., 2009. Is an environmental management system able toinfluence environmental and competitive performance? The case of the eco-management and audit scheme (EMAS) in the European Union. Journal ofCleaner Production 17 (16), 1444e1452.

Ismail, I.R., Hamid, R.A., Idris, F., 2012. PLS application in Journals of operationsmanagement: a review. In: Proceedings of Global Conference on Operations andSupply Chain Management, Indonesia, pp. 1e6.

Jabbour, A.B.L.S., Jabbour, C.J.C., 2009. Are supplier selection criteria going green?Case studies of company in Brazil. Industrial Management & Data Systems 109(4), 477e495.

Jabbour, C.J.C., 2010. In the eye of the storm: exploring the introductions of envi-ronmental issues in the production function in Brazilian companies. Interna-tional Journal of Production Research 48 (21), 6315e6339.

Jabbour, C.J.C., Santos, F.C.A., Nagano, M.S., 2010. Contributions of HRM throughoutthe stages of environmental management: methodological triangulationapplied to companies in Brazil. International Journal of Human ResourceManagement 21 (7), 1049e1089.

Jabbour, C.J.C., Silva, E.M., Paiva, E.L., Santos, F.C.A., 2012. Environmental manage-ment in Brazil: is it a completely competitive priority? Journal of CleanerProduction 21, 11e22.

Jackson, S.E., Renwick, D., Jabbour, C.J.C., Muller-Camen, M., 2011. State-of-the-artand future directions for green human resource management. Zeitschrift fürPersonalforschung (German Journal of Research in Human Resource Manage-ment) 25, 99e116.

Jackson, S.E., Schuler, R.S., 1995. Understanding human resource management in thecontext of organizations and their environments. Annual Review of Psychology46, 237e264.

Jacobs, B.W., Singhal, V.R., Subramanian, R., 2010. An empirical investigation ofenvironmental performance and the market value of the firm. Journal ofOperations Management 28 (5), 430e441.

Jamrog, J., Overholt, M.H., 2004. Building a strategic HR function: continuing theevolution. Human Resource Planning 27 (1), 51e63.

King, A.A., Lenox, M.J., 2001. Lean and green? An empirical examination of therelationship between lean production and environmental performance.Production and Operations Management 10 (3), 244e256.

Kline, R.B., 2005. Principles and Practice of Structural Equation Modeling. Guilford,New York.

Large, R.O., Thomsen, C.G., 2011. Drivers of green supply management performance:evidence from Germany. Journal of Purchasing & Supply Management 17,176e184.

Marcus, A., Fremeth, A., 2009. Green management matters regardless. Academy ofManagement Perspectives 23 (4), 17e26.

Maxwell, J., Briscoe, F., Schenk, B., Rothenber, S., 1998. Case study: Honda ofAmerican Manufacturing, Inc.: can lean production practices increase envi-ronmental performance? Environmental Quality Management 8 (1), 53e61autumn.

May, D.R., Flannery, L.B., 1995. Cutting waste with employee involvement teams.Business Horizons 38 (5), 28e38.

Molina-Azorin, J.F., Claver-Cortés, E., López-Gamero, M.D., Tarí, J.J., 2009. Greenmanagement and financial performance: a literature review. ManagementDecision 47 (7), 1080e1100.

Muller-Carmem, M., Jackson, S., Jabbour, C.J.C., Renwick, D., 2010. Green humanresource management. Zeitschrift für Personalforschung 24 (1), 95e96.

Murillo-Luna, J.L., Garcés-Ayerbe, C., Rivera-Torres, P., 2011. Barriers to the adoptionof proactive environmental strategies. Journal of Cleaner Production 19,1417e1425.

Ohno, T., 1988. Toyota Production System: Beyond Large Scale Production.Productivity Press, Cambridge.

Oliveira, P.W.S., Leone, R.G.J., 2008. Gestão estratégica de benefícios: proposição deummodelomatemático para assistênciamédica dos funcionários. RAMe Revistade Administração Mackenzie 9 (2), 104e127.

Osman, I., Ho, T.C.F., Galang, M.C., 2011. The relationship between human resourcepractices and firm performance: an empirical assessment of firms in Malaysia.Business Strategy Series 12 (1), 41e48.

Perron, G.M., Côte, R.P., Duffy, J.F., 2006. Improving environmental awarenesstraining in business. Journal of Cleaner Production 14 (6e7), 551e562.

Pestana, M.H., Gageiro, J.N., 2003. Análise de dados para ciências sociais: a com-plementaridade do SPSS. Sílabo, Lisboa.

Pettersen, J., 2009. Defining lean production: some conceptual and practical issues.The TQM Journal 21 (2), 127e142.

Pojasek, R.B., 2008. Framing your lean-to-green effort. Environmental QualityManagement 18 (1), 85e93. autumn.

Porter, M.E., Van Der Linde, C., 1995. Green and competitive: ending the stalemate.Harvard Business Review 73 (5), 120e134.

Renwick, D., Redman, T., Maguire, S., April 2008. Green HRM: A Review,Process Model, and A Research Agenda. University of Sheffield Manage-ment School, Sheffield, UK, Working Paper Series, Discussion Paper No.2008.1, pp. 1e32.

Robbins, S.P., Decenzo, D.A., 2004. Fundamentos de administração: conceitosessenciais e aplicações. Tradução Robert Brian Taylor, Revisão técnica ReinaldoO. da Silva, fourth ed. Pearson Prentice Hall, São Paulo.

Rothenberg, S., Pil, F.K., Maxwell, J., 2001. Lean, green, and the quest for superiorenvironmental performance. Production and Operations Management 10 (3),228e243.

Sarkis, J., 2001.Manufacturing’s role incorporateenvironmental sustainabilityeconcernsfor the newmillennium. International Journal of Operations & Production Manage-ment 21 (5e6), 666e686.

Sarkis, J., Gonzalez-Torre, P., Adenso-Diaz, B., 2010. Stakeholder pressure and theadoption of environmental practices: the mediating effect of training. Journal ofOperations Management 28, 163e176.

Shah, R., Ward, P.T., 2003. Lean manufacturing: context, practice bundles, andperformance. Journal of Operations Management 21, 129e149.

Simpson, D.F., Power, D.J., 2005. Use the supply relationship to develop lean andgreen suppliers. Supply Chain Management: An International Journal 10 (1),60e68.

Sosik, J.J., Kahai, S.S., Piovoso, M.J., 2009. Silver bullet or voodoo statistics?: a primerfor using least squares data analytic technique in group and organizationresearch. Group & Organization Management 35 (5), 5e36.

Sroufe, R., 2003. Effects of environmental management systems on environmentalmanagement practices and operations. Production and Operations Manage-ment 12 (3), 416e430.

Stoner, J.A.F., Freeman, R.E.,1999. Administração. Trad. Alice Alves Calado, fifth ed. LTC.

C.J.C. Jabbour et al. / Journal of Cleaner Production 47 (2013) 129e140140

Synodinos, N.E., 2003. The “art” of questionnaire construction: some importantconsiderations for manufacturing studies. Integrated Manufacturing Systems 14(3), 221e237.

Türk,K., Roolaht, T., 2007.Appraisalandcompensationof theacademic staff inEstonianpublic and private universities: a comparative analysis. Trames 11 (2), 206e222.

Vachon, S., Klassen, R.D., 2008. Environmental management and manufacturingperformance: the role of collaboration in the supply chain. International Journalof Production Economics 111 (2), 299e315.

Vais, A., Miron, V., Pedersen, M., Folke, J., 2006. “Lean and green” at a Romaniansecondary tissue paper and board mill e putting theory into practice. ResourcesConservation & Recycling 46, 44e74.

Voorde, K.V.D., Paauwe, P., Veldhoven, M.V., 2010. Predicting business unit perfor-mance using employee surveys: monitoring HRM-related changes. HumanResource Management Journal 20 (1), 44e63.

Wagner, M., 2007. Integration of environmental management with other manage-rial functions of the firm: empirical effects on drivers of economic performance.Long Range Planning 40 (6), 611e628.

Wetzels, M., Odekerken-Schroder, G., Oppen, C.V., 2009. Using PLS path modelingfor assessing hierarchical construct models: guidelines and empirical illustra-tion. MIS Quarterly 33 (1), 177e195.

Wilkinson, A., Hill, M., Gollan, P., 2001. The sustainability debate. InternationalJournal of Operations & Production Management 21 (12), 1492e1502.

Womack, P.J., Jones, D.T., Roos, D., 2004. A máquina que mudou o mundo. Elsevier,Rio de Janeiro.

Woolverton, A., Dimitri, C., 2010. Green marketing: are environmental and socialobjectives compatible with profit maximization? Renewable Agriculture andFood Systems 25 (2), 90e98.

Yang, C., Lin, S., Chan, Y., Sheu, C., 2010. Mediated effect of environmentalmanagement on manufacturing competitiveness: an empirical study. Interna-tional Journal of Production Economics 123, 210e220.

Yang, M.G., Hong, P., Modi, S.B., 2011. Impact of lean manufacturing and envi-ronmental management on business performance: an empirical study ofmanufacturing firms. International Journal of Production Economics 129 (2),251e261.