ly

Sermamarkester

Keywords:Literature reviewOR in environment and climate changeOR in societal problem analysisOR in sustainability

ations research methods and related models, i.e. formal modeling, for closed-loop SCM and reverse

themanagement of physical, logical, and nancial ows in networksof intra- and inter-organizational relationships jointly adding valueand achieving customer satisfaction (Mentzer et al., 2001; Stock &Boyer, 2009). From a process-oriented or cross-functional perspec-tive, SCMcomprisesplanning, sourcing, production, anddistribution

emically, swthof reland especial

the past decade (Min&Kim, 2012; Seuring&Mller, 2008a). Ition to a large variety of empirical research papers that utiliresearch, case study, and broad-based empirical surveys, numerouspublications employ formal, mathematical models for practice andtheory-driven research. Models are a simplied representation orabstraction of reality, and related research differentiates betweenconceptual models dened as a set of concepts suitable to representbut not explain real-life objects or processes and quantitativemodels that are based on a set of variables and their causal relation-ship (Bertrand & Fransoo, 2002; Meredith, 1993).

Corresponding author. Address: Supply Chain Management, Faculty of Businessand Economics, University of Kassel, Untere Knigsstr. 71, D-34117 Kassel,Germany. Tel.: +49 561 804 7517.

European Journal of Operational Research 233 (2014) 299312

Contents lists availab

European Journal of O

.eE-mail address: brandenb@uni-kassel.de (M. Brandenburg).ationsmanagement using the supply chain (SC) perspective throughthe incorporation of upstream (input) and downstream partners(output) into the boundary of investigation and management(Bettley & Burnley, 2008). Traditionally, SCM has been dened as

and product recovery.The increasing importance of this eld, acad

andeconomically, is reectedby thegeometric groentic publications during the past two decades a0377-2217/$ - see front matter 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.ejor.2013.09.032ocially,ted sci-ly so inn addi-ze eldThe integration of environmental and social aspects witheconomic considerations, known as the triple-bottom-line (TBL)dimensions of organizational sustainability (Elkington, 1998,2004), has continuously gained relevance for managerial decisionmaking in general and for supply chain management (SCM) (Carter& Rogers, 2008) and operations management (Drake & Spinler,2013; Kleindorfer, Singhal, & van Wassenhove, 2005) in particular.

Organizations have rethought and redened the concept of oper-

In contrast to traditional SCM, which typically focuses on eco-nomic and nancial business performance, sustainable SCM(SSCM) is characterized by explicit integration of environmentalor social objectives which extend the economic dimension to theTBL (Seuring & Mller, 2008a). In this context, SSCM focuses onthe forward SC only (Seuring & Mller, 2008a) and is comple-mented by closed-loop SCM (CLSCM) (Guide & van Wassenhove,2009; Lebreton, 2007) including reverse logistics, remanufacturing,Supply chain managementSustainability

1. Introductionlogistics has been effectively reviewed in previously published research. This situation is in contrast tothe understanding and review of mathematical models that focus on environmental or social factors inforward supply chains (SC), which has seen less investigation. To evaluate developments and directionsof this research area, this paper provides a content analysis of 134 carefully identied papers on quanti-tative, formal models that address sustainability aspects in the forward SC. It was found that a prepon-derance of the publications and models appeared in a limited set of six journals, and most wereanalytically based with a focus on multiple criteria decision making. The tools most often used comprisethe analytical hierarchy process or its close relative, the analytical network process, as well as life cycleanalysis. Conclusions are drawn showing that numerous possibilities and insights can be gained fromexpanding the types of tools and factors considered in formal modeling efforts.

2013 Elsevier B.V. All rights reserved.

logistics (Supply-Chain Council, 2008) but is not exclusively focusedon one of these areas (Cooper, Lambert, & Pagh, 1997).Article history:Available online 4 October 2013

Sustainability, the consideration of environmental factors and social aspects, in supply chain manage-ment (SCM) has become a highly relevant topic for researchers and practitioners. The application of oper-Quantitative models for sustainable suppDevelopments and directions

Marcus Brandenburg a,d,, Kannan Govindan b, Josepha Supply Chain Management, Faculty of Business and Economics, University of Kassel, GbDepartment of Business and Economics, University of Southern Denmark, Odense, DencDepartment of Management, School of Business, Worcester Polytechnic Institute, WorcdDepartment of Production Management, Technical University of Berlin, Germany

a r t i c l e i n f o a b s t r a c t

journal homepage: wwwchain management:

arkis c, Stefan Seuring a

ny

, MA, USA

le at ScienceDirect

perational Research

l sevier .com/locate /e jor

and Seuring, 2012), many of which are quite recent in develop-ment. A comprehensive review of these models is not currently

of Oavailable and thus it is timely to take an inventory of the research.The lack of a comprehensive understanding of modeling-basedSSCM research is surprising since the non-sustainability modelingeld has a well-developed traditional research focus on forwardSCM. It may be that research focusing on CLSCM has caused manymodeling researchers to overlook this forward SCM eld in contextto sustainability.

To help further catalyze research in this area, which hasnumerous opportunities to improve organizational, industrial,and commercial sustainability, further understanding of thecommon and unique modeling characteristics is needed. SomeSSCM reviews currently exist, but most of these reviews aredescriptive (e.g. Carter & Rogers, 2008; Fleischmann et al.,1997; Seuring & Mller, 2008a). Although somewhat descrip-tive, this paper provides additional insightful discussion,analyzing a number of important eld advancing questions asdiscussed below.

Which aspects and factors are considered in existing quantita-tive SSCM models? What are the limits of these models and whatissues remain? What feasible and fruitful opportunities forfurther research exist? To help understand the history anddirection of SSCM modeling efforts and to answer these ques-tions, this paper presents a content analysis (Krippendorff,1980; Mayring, 2002, 2008) of related literature to assess recentdevelopments and future directions of quantitative, formal mod-eling in the SSCM context. The rich descriptions offered and theoverall lines of research identied this way, often have tremen-dous impact on future research. Therefore, we also discussoverarching lines of research as well as gaps and future researchdirections.

The remainder of this paper is structured as follows. In the nextsection, a brief overview of related literature reviews on SSCM isgiven. The subsequent section describes the methodology appliedin this paper and is followed by a representation of the resultsobtained by the content analysis. Remarks comprising the summa-rized ndings and a related discussion, limitations, and future re-search perspectives conclude this paper.

2. Insights from previous literature reviews

To justify the need for the content analysis presented in this pa-per and to position its results to extant scientic research, formerreviews of scientic literature on SSCM are summarized. Existingliterature reviews on SSCM can be categorized into reviews pub-lished prior to 2008 and recent reviews published within the lastve years. The purpose of this background on previous literaturereviews is to help derive relevant information and structures forthis study. The background literature also helps to identify open is-sues in model-based SSCM research. These recent reviews are as-For CLSCM, quantitative models are often applied and practical(Fleischmann et al., 1997; Srivastava, 2007; Sasikumar & Kannan,2008a, 2008b, 2009). In contrast to this circumstance, the majorityof models employed for SSCM are more conceptual. Only about oneout of nine papers on SSCM utilizes formal models (Seuring & Ml-ler, 2008a). In recent years, the quantity of formal modeling effortshas started to increase.

It is evident from literature that (reverse-oriented) CLSCMmod-els are more popular (Ilgin & Gupta, 2010; Min & Kim, 2012), but asignicant number of (forward) SSCM models do exist (Hassini,Surti, & Searcy, 2012; Min & Kim, 2012; Seuring & Mller, 2008a;

300 M. Brandenburg et al. / European Journalsessed with regards to SCM perspectives, e.g. level and actor ofanalysis, sustainability, i.e. the dimensions of the TBL, andresearch designs.2.1. Literature reviews prior to 2008

The earliest related literature reviews (Gungor & Gupta, 1999;Kleindorfer et al., 2005) identify green product and process devel-opment, green operations management, remanufacturing, andCLSCM as areas to integrate planet- and people-related issues intoSCM, but the reviews do not include social aspects of SSCM.Bloemhof-Ruwaard, van Beek, Hordijk, and van Wassenhove(1995) focused on operations research (OR) applications in thecontext of environmental management (EM) and suggest a con-ceptual SC-EM-framework. Daniel, Diakoulaki, and Pappis(1997) apply this framework in their survey of OR-related envi-ronmental planning and categorize related OR methods intodescriptive approaches for observation and analysis and norma-tive methods for solution identication. ReVelle (2000) providesan overview on the application of OR methods for the manage-ment of water resources, solid waste, and air quality and outlinesdifferent normative models for these areas. Sbihi and Eglese(2007) focused specically on combinatorial optimization prob-lems in green logistics, which comprises reverse logistics, wastemanagement, and vehicle routing and scheduling. While theseearly published reviews paved the way for SSCM research, theyare not able to inform on current developments and future trendsof related model-based research.

2.2. Literature reviews after 2008

Recent reviews of SSCM literature can be categorized as eithergeneral or focused on empirical research or quantitative modelsand metrics. Table 1 overviews 14 recent reviews regarding theirresearch focus and characteristics, such as time horizon, numberof reviewed papers, main journals, employment of keyword searchand content analysis as well as taken perspectives on SCM andsustainability.

In contrast to reverse logistics or remanufacturing, OR meth-odologies and analytic approaches for forward SSCM play asubordinate role in the published research (Ilgin & Gupta,2010; Min & Kim, 2012). As shown in Table 2, approximatelyonly one out of ten SSCM papers employs a research methodwhich is based on quantitative models using formal OR model-ing techniques.

With regards to the SCM perspective, extant research showsthat sustainability is often externally motivated by government,customers, or stakeholders (Gold, Seuring, & Beske, 2010a,2010b; Seuring & Mller, 2008a). The literature also shows that avertical coordination and a SC-wide implementation are required(Carter & Rogers, 2008). In contrast to this focus, empirical researchon SSCM mainly focuses on single rms and neglects inter-organi-zational aspects (Carter & Easton, 2011). This conict leads to thequestion of whether model-based research takes into account theintercompany perspective and if the role and inuence of legalauthorities or other stakeholders is adequately reected in quanti-tative SSCM models. Furthermore, Hassini et al. (2012) show thatsustainability metrics are most often designed for manufacturingrms. Hence it should be assessed which SC sectors are in the focusof model-based SSCM research.

Holistic approaches of SSCM that reect all three sustainabilitydimensions are relatively rare in the academic literature (Seuring& Mller, 2008a). However, empirical research shows the growingrelevance of multiple sustainability dimensions (Carter & Easton,2011). Given that SSCM can positively inuence a rms protabil-ity, performance, and competitive advantage (Carter & Rogers,2008; Gold et al., 2010b; Golicic & Smith, 2013), SSCM research

perational Research 233 (2014) 299312tends to focus primarily on environmental issues (Seuring &Mller, 2008a), while social facets are widely neglected in empirical(Gold et al., 2010a) and in analytical SSCM modeling research

Main journalsa Keyword search content

Perspective on SCM sustainability

n.JC

JCEJ

JCn.TRn.

n.n.

n.

n.

IJPJCIJPn.

of OpTable 1Recent reviews on SSCM literature.

Research focus Author(s) and year Timehorizon

Number ofreviewedpapers

General Carter and Rogers (2008) n.a. n.a.Seuring and Mller(2008a)

19942007 191

Min and Kim (2012) 19952010 519

Empirical Gold et al. (2010a) 19942007 70Gold et al. (2010b) 19942007 70Carter and Easton (2011) 19912010 80Sarkis, Zhu, and Lai(2011)

19952010 150

Sarkis (2012) 20002010. 100Golicic and Smith (2013) 20002011 77

Quantitativemodels andmetrics

Ilgin and Gupta (2010) 19992010 540

Dekker et al. (2012) n.a. 60

Hassini et al. (2012) 20002010 87Seuring (2012) 19942010 36

Tang and Zhou (2012) n.a. 56

a Not including journals with a pure empirical focus.b Only rough categorization based on very few categories and dimensions.

M. Brandenburg et al. / European Journal(Tang & Zhou, 2012). Tang and Zhou (2012) observe that environ-mental factors in quantitativemodels mainly include the consump-tion of natural resources and the emission of waste and pollution,while social aspects are related to only customers and producers.Assessing the literatures usage of the three sustainability dimen-sions in greater detail, e.g. which metrics are suitable to representsustainability factors in formal SSCM models and which perspec-tives are taken in holistic SSCM models, would identify what ave-nues exist to further integrate holistic TBL measures and theresulting performance impacts into SSCM formal modeling.

In contrast to SSCM research in general, which focuses on winwin approaches to sustainability (Seuring & Mller, 2008a), formalmodeling research is dominated by trade-off based modeling ap-proaches (Seuring, 2012). An assessment of the main purposes(descriptive or normative) of SSCM models still requires investiga-tion. Although a lack of multi-criteria decision making (MCDM) ap-proaches for green logistics seems to exist (Dekker, Bloemhof, &Mallidis, 2012), preferred types and techniques for forward SSCMmodels have not been identied and analyzed. This lack of identi-cation of prevalent modeling approaches is in contrast to Ilginsand Guptas (2010) ndings for green (re-)manufacturing andproduct recovery, where discrete-event simulation (DES), fuzzylogic, genetic algorithms (GA), and mixed-integer linear program-ming (MILP) are identied as preferred modeling techniques.

The research design is important to consider in the literaturereview because it can help identify the current situation in the eldand how it may be advanced by varying research methodology.Previous research has found that empirical SSCM studies are more

Table 2Relevance of forward SSCM formal models in scientic research.

Author(s) and year Size of initial paper sample

Seuring and Mller (2008a) 191Min and Kim (2012) 519Hassini et al. (2012) 707Seuring (2012) 306

Note: The remaining literature reviews did not contain information reanalysis

a. No no Undisclosed TBLLP, POM, IJPR Yes yes Forward TBL

LP, IJPR, IJPE,OR

Yes yesb Forward and reverse TBL

LP Yes yes Forward TBLa. Yes yes Forward TBLE No no Undisclosed TBLa. No no Forward and reverse environmental

a. No no Forward and reverse environmentala. Yes no Undisclosed environmental

a. No yesb Forward and reverse environmental

a. No no Forward and reverse logistics environmental

R, IJPE, EJOR Yes yesb Forward and reverse TBLLP, EJOR, IJPE,R

Yes yes Forward TBL

a. No no Forward and reverse TBLerational Research 233 (2014) 299312 301prevalent than conceptual or formal modeling research designs(Seuring & Mller, 2008a). It has also been found that forwardSSCM research can benet from OR applications, ideally combinedwith rigorous empirical studies (Min & Kim, 2012; Seuring, 2012).But the ndings in these papers have only started to identify theseconcerns, and further investigation can provide additional insights.For example, a sectoral focus of scientic SSCM research is of par-ticular interest, which is not yet covered in model-based SSCM re-search. Empirical studies focus on transportation, textile, andconsumer products sectors while the automotive, chemical, andelectronics industries have fewer investigations (Carter & Easton,2011; Gold et al., 2010a, 2010b). Again, determining whether mod-el-based SSCM research considers similar sectoral emphasis wouldprovide insight into what sectors are underrepresented and alsoshed light on why certain sectors might be overrepresented. A po-tential mismatch can be predicted by the fact that papers on SSCMmetrics preferably deal with manufacturing sectors such as auto-motive or electronics industries (Hassini et al., 2012). Identifyinga sectoral preference or lack thereof can provide guidance to policymakers and researchers on what sectors need further academic andpolicy intensive modeling research.

2.3. Research questions

To assess research developments and directions for formalmodeling in forward SSCM, we ask the following researchquestions:

Number with SSCM formal models Share (%)

21 1146 987 1236 12

quired for these calculations.

lyzed according to the structural dimensions and ana-

of Olytic categories to identify relevant issues and tointerpret the results.

3.2. Material collection

This literature review is bounded to only include: (a) scienticresearch from the last 15 years; (b) formal models; (c) SSCM; and(d) forward SC. Hence, each reviewed paper had to match fourltering criteria:

(a) The manuscript must be written in English language andpublished in peer-reviewed journals between 1994 and2012.

(b) Empirical manuscripts using statistical approaches for eval-uating causal relationships were excluded from the analysis.

(c) Publications on ethical behavior of purchasers (e.g. corrup-tion) or with a non-managerial focus (e.g. technical or polit-ical science) were excluded from the analysis.

(d) Papers focusing on reverse logistics, remanufacturing, orCLSCM were not considered.1. Which SCM dimensions exist with formal models for SSCM?2. Which sustainability aspects of forward SC are reected in these

models?3. Which models and tools are employed in quantitative SSCM

research?4. What research designs exist and which future directions should

be addressed?

Sustainability aspects within one or more industry sectors, e.g.CO2 emission of the manufacturing industry, or within geograph-ical regions or macro-economies, such as rural, poor areas ordeveloping countries, are included in our analysis. The papersfocus is on formal, quantitative models capable of predictingthe outcome of actions or theoretically evaluating variousdynamic properties of complex problems (Mikkola, 2005). In theremainder of this paper, the terms formal and quantitativewill be used interchangeably for models that include (descriptiveand normative) OR methods but exclude statistical approaches,e.g. regression or structural equation models, to evaluate empiri-cal data.

3. Methodology

3.1. Content analysis

To address the four research questions, a thorough review of pa-pers on quantitative models for SSCM is performed. To ensure therequired methodological rigor, this literature review employs thesystematic process of content analysis (Krippendorff, 1980; LageJunior & Godinho Filho, 2010) that consists of four iteratively exe-cuted steps (Mayring, 2002, 2008):

Step 1. Material collection: The material to be collected and theunit of analysis are dened and delimited.

Step 2. Descriptive analysis: Formal aspects of the material areassessed.

Step 3. Category selection: Structural dimensions including themajor topics of analysis and related analytic categorieswith detailed classications of each structural dimen-sion are selected to be applied to the collected material.

Step 4. Material evaluation: The content of the papers is ana-

302 M. Brandenburg et al. / European JournalThe most common way of acquiring the publications sample isthrough a keyword-based search using electronic databases and li-brary services (Seuring & Gold, 2012). This approach is especiallyrecommended for covering a specic topic that can be broadly ad-dressed. Complementary, the paper search can be focused on se-lected journals. This allows employing a broader search stringand thus makes it easier to assess all related papers on a certain to-pic, although some relevant papers in other journals might bemissed (Seuring & Gold, 2012). These keyword-based searchescan be complemented by cross-referencing for further relevantpublications (Athanasopoulou, 2009) or by employing bibliometricsoftware (Linnenluecke & Grifths, 2012). These complementarysearch strategies were combined, resulting in four steps of papercollection applied in this paper.

In a rst step, all papers reviewed by Seuring (2012, 36 papers)and Hassini et al. (2012, 87 papers) were considered. This reectsbroad searches in databases (Emerald, Elsevier,Wiley, Springer, Ebs-co, Scopus, Metapress) which were performed by these authors andlimited by them to particular keyword combinations (e.g. (sustain-able OR green) AND (supply AND chain)). Fifty-six papers reviewedby Tang and Zhou (2012) were additionally taken into account.Forty-six out of the total 179 papersmatched the four ltering crite-ria for this study andwere the initial sample set. Hassini et al. (2012),Seuring (2012) andTang andZhou (2012) identiedDecision SupportSystems (DSS), European Journal of Operational Research (EJOR), Inter-national Journal of ProductionEconomics (IJPE), International Journal ofProduction Research (IJPR), Journal of Cleaner Production (JCLP), andTransportation Research Part E (TRE) as relevant journals for quanti-tative SSCM models (see Table 1 in Section 2.2 or Table 6 in Sec-tion 4.1 for the distribution of the 46 papers over these journals).Two reviews published in 2012were not considered in this rst stepdue to an undisclosed paper sample (Min & Kim, 2012) or a func-tional and purely environmental focus (Dekker et al., 2012).

In a second step, a journal-specic search was performed whichwas limited to the six journals using the broader search stringsustain in title, abstract, and keyword elds. This journal-specic search resulted in 1016 papers out of which an additional81 papers matched the four ltering criteria and hence wereselected for the review.

In a third step, the paper sample was completed by cross-refer-encing using ReVelle (2000, 39 papers), Radulescu, Radulescu, andRadulescu (2009, 30 papers), and Seuring (2012, 8 papers). Six pa-pers out of these 77 cited manuscripts matched the four lteringcriteria and hence were selected for the review.

In a fourth step, the literature sample was validated by employ-ing the bibliometric software HistCiteTM (version 12.03.17) (Gar-eld, 2004). With this software, 25 manuscripts were identiedthat were cited by at least ve of the selected papers. All but oneof these 25 manuscripts describe a particular modeling method(e.g. Saaty, 1980, 10 citations from the paper sample) or literaturereviews on SSCM models (e.g. Bloemhof-Ruwaard et al., 1995, 9citations) or were already included in the paper sample (e.g. Sheu,Chou, & Hu, 2005, 8 citations). One additional paper (Handeld,Walton, Sroufe, & Melnyk, 2002, 7 citations) from these 25manuscripts, which was not identied by the rst three data col-lection steps, matched the four ltering criteria and thus wasadded to the paper sample.

In total, approximately 1400 papers were considered, out ofwhich 134 were found to meet the criteria of this study. This selec-tion ratio matches the 10% occurrence rate for modeling paperswhich was mentioned in Section 2.2. Table 3 illustrates the resultsof the paper collection process and the paper validation test.

3.3. Criteria for the descriptive analysis

perational Research 233 (2014) 299312The temporal distribution of papers over the study time horizonis assessed. For a more meaningful descriptive analysis, the distri-bution of manuscripts from the sample over time is compared to

the overall growth trend of publications in SCM and modeling jour-nals, as indicated by papers issued in EJOR and in IJPE, which are

analysis. These analytic categories listed in Table 4 are deriveddeductively, before the material is analyzed, and inductively,

Table 3Results of the paper collection process.

Collection method Source # of papers Relevant worksc

Selected from other reviews Hassini et al. (2012) 87 7Seuring (2012) 36 35Tang and Zhou (2012) 56 4

Journal-specic keyword search DSS 29 3EJOR 80 13IJPE 75 20IJPR 125a 12JCLP 825 31TRE 7 2

Cross-referencing Other papersb 77 6Bibliometric software HistCiteTM 25 1

Total 1422d 134

a Split of search criteria (in title, in abstract, in keywords).b ReVelle (2000, 39 papers), Radulescu et al. (2009, 30 papers), Seuring (2012, 8 papers).c Each paper is assigned only once to exclude double counting.d Not adjusted for papers found several times by different search strategies.

M. Brandenburg et al. / European Journal of Operational Research 233 (2014) 299312 303chosen as related journal proxies. In order to compare develop-ments in empirical, model-based and general SSCM research, thetemporal distribution of the paper sample is compared to the pa-pers of Seuring and Mllers (2008a) general review and to theempirical papers reviewed by Gold et al. (2010a, 2010b). Further-more, the distribution of papers across journals is analyzed. Toavoid bias resulting from a journal-specic paper search, this anal-ysis also shows the distribution of the 46 papers selected from Has-sini et al. (2012), Seuring (2012), and Tang and Zhou (2012).Additionally, the descriptive analysis provides information on geo-graphical position of the contributing author afliations (academicinstitutions), about inuential research institutions (regarding thenumber of citations) and the citation impact of the reviewed papersample. The HistCiteTM program was employed for the bibliometriccitation analysis (Gareld, 2004).

3.4. Category selection

Corresponding to the four research questions of this study, fourstructural dimensions SCM, sustainability, modeling, and re-search directions were dened and grouped by categories for this

Table 4Structural dimensions and analytic categories.Structuraldimension

Analytic categories (in alphabetical order)

SCM Primary actor ofanalysis

Carrier, distributor, industry/macro-economy

Level of analysis Chain, dyad, rm, function, industry, macro-Process of analysis Deliver, make, plan, return, sourceFunctionalapplication area

Construction project, logistics, network desigsourcing, SCM, information technology, tech

Sustainability Economic, economic-environmental, environ

Modeling Model purpose Descriptive deterministic, descriptive stoModel type Analytical, heuristics, hybrid, mathematicalModeling technique Articial intelligence, business game, discret

making, multi objective, simple heuristics, sSolution approach Analytic hierarchy process/analytic network

reasoning, data envelopment analysis (DEA),programming, greedy randomized adaptiveprogramming/mixed integer linear programmparticle swarm optimization, queuing, rough

Research Observed industry Agriculture, apparel, automotive, bicycle, biofurniture, health care, information technolog

Numerical analysis Empirical data, generic example, noneSuggestedperspective

Extend/validate, none, specicdeveloped from the analyzed material by means of generalization(Mayring, 2008; Seuring & Mller, 2008a).

The SCM dimensions are from a literature review presented byHalldorsson and Arlbjrn (2005) and from the Supply Chain Oper-ations Reference (SCOR) model (Supply-Chain Council, 2008). Thesustainability categories are based on variations of the TBL (Carter& Rogers, 2008). The modeling dimension is assessed based onmodel purpose and type. The model purpose distinguishes be-tween the analytic categories as dened by Shapiro (2007). Themodel type keywords were quite extensive and linked to a multi-tude of tools and techniques and the employed solution ap-proaches. The resulting category system, which is depicted inFig. 1, is based on the classication of Kleijnen (2005) and Sasiku-mar and Kannan (2009). In the research design dimension, theassessment is based on the industrial sector, the data basis of thepresented applications, and the research perspective such aswhether or not there is an extension.

To facilitate an exhaustive categorization of each paper, theanalytic categories are supplemented with various/other andnot applicable categories. Furthermore, the assignment of papersto analytic categories is unique for each structural dimension., legal authority, manufacturer, retailer, warehousing, wholesaler

economy, network

n, outsourcing/offshoring, planning, pricing, product development, production,nology, waste management

mental, holistic, social, socio-economic, socio-environmental

chastic, normative deterministic, normative stochasticprogramming, simulatione-event simulation (DES), game theory, meta-heuristics, multi-criteria decisioningle objective, spreadsheet calculation, system dynamics, systemic modelsprocess (AHP/ANP), ant colony optimization, Bayesian networks, case baseddifferential evolution, dynamic programming, fuzzy logic, genetic algorithm, goalsearch procedure, inputoutput-analysis (IOA), life cycle analysis (LCA), linearing (LP/MILP), metrics, neural networks, nonlinear programming, petri net,set, simulated annealing, variation inequality

technology, chemical/ pharmaceutical, electronics, energy, food & beverages,y, macro-economy, metal, paper, retail, transportation, utilities

304 M. Brandenburg et al. / European Journal of O3.5. Methodological rigor

The required methodological rigor of this analysis is ensured inall steps of the content analysis (see Seuring & Gold, 2012, for fur-ther details). A replicable and hence reliable material collection isachieved by focusing both on papers that were assessed by threerecent reviews on SSCM (Hassini et al., 2012; Seuring, 2012; Tang& Zhou, 2012) and on papers that are obtained by a keyword-basedsearch in specic journals which were identied as most relevantby these three recent reviews. Internal validity is achieved byensuring that the paper coding is performed by at least tworesearchers, thereby also ensuring inter-coder reliability. Resolvingdifferences in the coding and reaching an agreement on how eachaspect would be coded was the approach of choice. Most of the se-lected structural dimensions and analytic categories were taken upin a deductive approach, which ensures construct validity as this isbased on respective literature (as described in Section 3.4). Thiswas complemented by a few categories selected inductively. Tostrengthen external validity and rigor of the material evaluation,intermediate results of this analysis were presented to and dis-cussed with scientic audiences at ve international conferences.1

Overall, the research process is documented in a transparent mannercontributing to its objectivity.

Fig. 1. Analytic categories of the structural dimension

1 4th World P&OM Conference and 19th International Annual EurOMA conference(July 2012, Amsterdam, Netherlands), 2nd International Workshop on Eco-EfcientBased Green SCM, (October 2012, Odense, Denmark), sessions of the GermanAcademic Association for Business Research scientic commissions OperationsResearch (January 2013, Wuppertal, Germany) and Sustainability Management(September 2012, Hamburg, Germany), workshop of the Gesellschaft fr OperationsResearch e. V. (March 2012, Goslar, Germany).perational Research 233 (2014) 2993124. Results

4.1. Descriptive analysis

Overall it can be stated that model-based SSCM is a comparablyyoung and increasingly developing research discipline. This eldsrobust growth is illustrated by the temporal distribution of thisstudys sample of 134 papers when compared to the growthdistribution of the overall set of publications within EJOR and IJPE(Fig. 2). Also, the temporal distribution of this studys paper sampleis compared to the distribution of the 191 manuscripts of Seuringand Mllers (2008a) study (Table 5).

Since 1994, the number of analytical SSCM model papers showsa stronger growth (+24.5% compound annual growth rate (CAGR))than the overall number of papers published in EJOR and IJPE(+12.6% CAGR). Fig. 2, which depicts the annual share of publishedpapers and the number of SSCM papers on the ordinate axes, illus-trates these trends in greater detail. In the rst decade of thestudys time horizon, the annual share of published SSCM analyti-cal modeling papers was lower (below 3.0%) than the respectiveshare of papers issued in EJOR and IJPE (between 3.7% and 4.4%),which indicates the relatively smaller relevance of SSCM modelingin scientic research during that time. This result was reversed in2003 for the rst time (4.5% annual share of SSCMmodeling papersvs. 4.2% annual share of papers published in EJOR and IJPE) and hascontinuously seen stronger shares since 2007 (on average an 11.7%annual share of SSCM modeling papers vs. a 7.2% annual share ofpapers published in EJOR and IJPE). More than half of the SSCMmodeling papers (82 manuscripts) were published within the lastve years of the considered time horizon.

Modeling (based on Sasikumar & Kannan, 2009).

of OpM. Brandenburg et al. / European JournalUsing Seuring and Mllers (2008a) sample illustrates that theoverall eld of SSCM research was strong before 2007 while SSCMmodeling papers published during this time were relativelyuncommon. A comparison to the paper sample reviewed by Goldet al. (2010a, 2010b) leads to similar observations. These results,as depicted in Table 5, indicate that conceptual and empiricalmethods paved the way for a model-based prescriptive SSCMresearch.

The distribution of papers over journals as depicted in Table 6shows that nearly two-thirds of the 46 papers chosen from Hassiniet al. (2012), Seuring (2012), and Tang and Zhou (2012) in the rstpaper collection step were published in one of the six journals(DSS, EJOR, IJPE, IJPR, JCLP, TRE) which then were considered inthe journal-specic keyword search. Beyond this, Table 6 depictsthe overall distribution of all 134 papers over journals.

The analysis of the geographical position of the contributingauthors institutions reveals that universities from Europe (61papers), North America (44 papers), or Asia (41 papers) representthe vast majority of publications. Australian (6 papers), LatinAmerican (4 papers), and African (3 papers) research publications

Fig. 2. Paper sample compared to

Table 5Temporal distribution of general, empirical, and model-based SSCM papers (19942007).

Year 1994 1995 1996 1997 1998 1999 2000

Gen.a 2 3 5 10 12 4 14Emp.b 1 0 2 3 6 1 6Mod.c 1 2 0 2 4 3 2

a Papers reviewed by Seuring and Mller (2008a).b Papers reviewed by Gold et al. (2010a, 2010b).c SSCM modeling papers from our paper sample.

Table 6Distribution of papers over journals.

Journal JCLP IJPR IJPE TRE

Step 1 9 7 4 4Subtotal 28 (= 61%)All steps 40 19 24 6

a 16 other journals with one paper each.b 18 other journals with one paper each.erational Research 233 (2014) 299312 305are severely underrepresented in this research area. Only 21 papersrepresent collaborative inter-continental research. These manu-scripts mainly stem from research cooperation of North Americanuniversities (15 papers) with institutions from other continents(Asian: 8 papers, European: 5 papers, Australian and Latin Ameri-can: 1 paper each).

A citation analysis shows that 2049 publications refer to at leastone of the reviewed papers, i.e. on average each paper of the sam-ple has more than 15 citations in Thomson Reuters Web of Sci-ence

while the median is 8 citations. Fifty-eight papers of the

sample have at least 10 citations in total, and 80 papers are citedat least once a year, while only 34 papers of the sample are notyet cited. These gures indicate the overall scientic relevance ofthe reviewed paper sample. However, the paper sample does notshow a strong internal coherence with regards to citations; only31 papers are cited by another paper of the sample. This indicatesthat the scientic eld of model-based SSCM research is still scat-tered, and major streams of thought have not yet developed. Fur-thermore, the citation analysis identies that ve of the mostinuential institutions are located in the USA: Clark University

SCM and modeling papers.

2001 2002 2003 2004 2005 2006 2007

21 16 20 17 25 12 307 4 11 5 10 3 112 3 6 4 8 4 11

EJOR DSS OR Others Total

3 1 2 161a 4618 (= 39%) 46

20 5 2 181b 134

inter-organizational perspective on a dyad, chain, or network (38papers) or macroscopic views on an industrial sector or a macro-

ere

of Oeconomy (25 papers).In 42 papers SSCM models are not limited to a particular SCOR

process. The majority of models support some planning processes(325 citations), Michigan State University (184 citations), BostonCollege, Emory University, North Carolina State University (165citations each).

4.2. Evaluating the status of research

4.2.1. SCM dimensionThe SCM dimensions, as shown in Table 7, include four major

characteristics: (1) the primary actor that is the focus of the study;(2) the organizational level of analysis ranging from internal func-tions to broad macro-economic focus; (3) the SCOR process cov-ered in the model, and (4) the functional application area (e.g.SCM, logistics, or product development).

In SSCM models, manufacturing companies dominate as pri-mary actors of analysis (67 papers), while carriers (Lee, Dong, &Bian, 2010; Lovric, Li, & Vervest, 2012), distributors (Zanoni & Zav-anella, 2011) and retailers (Edwards, McKinnon, & Cullinane, 2010)are seldom the focus of these studies. A considerable number of pa-pers analyze industry sectors (15 papers) and legal authorities (11papers) or remain unspecic (18 papers various, 19 papers n.a.) on the primary actors. The level of analysis shows a clear pref-erence for the intra-organizational models that focus on a specicfunction or rm (66 papers) when compared to models that take an

Table 7Evaluation results structural dimension SCM.

Primary actor of analysis Level of analysis

Carrier 2 Function 27Distributor 1 Firm 39Ind./macro-econ. 15 Dyad 7Legal authority 11 Chain 6Manufacturer 67 Network 25Retailer 1 Industry 17Warehousing 0 Macro-econ. 8Wholesaler 0 Various/other 1Various/other 18 n.a. 4n.a. 19

a Construction project, pricing, information technology, and waste management w

306 M. Brandenburg et al. / European Journal(53 papers). Yet, sourcing (12 papers), transformation (18 papers),or delivery processes (9 papers) are less often modeled. None of theselected papers focus on the return stage, but that is due to ourelimination of reverse logistics oriented papers from our analysis.

Functionally, SSCM modeling research targets production (40papers) or general SCM (16 papers). Manufacturing-related paperscan be found in nearly every year of the study horizon, while nineof the general SCM papers were published within the last veyears. In the nine logistics-related papers as well as in the 11sourcing-oriented ones, the intra-organizational and the inter-organizational perspectives on the level of analysis are evenly dis-tributed. Network design models (13 papers) seem to represent anew trend in SSCM research, because only the publication of Agrell,Stam, and Fischer (2004) is older than three years. Issues of prod-uct design are considered in nine papers which in seven cases dealwith a manufacturer as a focal company. Whitefoot and Skerlos(2012), who elaborate on design incentives from a legal authorityperspective, and Andersson, Hogaas Eide, Lundqvist, and Mattsson(1998), who integrate sustainability aspects in product develop-ment from an industry point of view, are the two exceptions.4.2.2. Modeling dimensionIn the evaluation of this dimension we consider the three levels

of modeling from Fig. 2, but also a general purpose-environmentcharacteristic is added (see Table 8). The purpose-environment fo-cuses on the type of research for which normative (prescriptive,problem solving oriented) or descriptive (seeking to evaluate orunderstand a phenomenon) models are used. The assumed envi-ronment may include a deterministic or stochastic approach. Thusfour categories (descriptivedeterministic, etc.) exist for the pur-pose-environment evaluation of the modeling dimension. Theremaining three dimensions in Table 8 are directly linked to themodeling levels in Fig. 1.

Normative SSCM models (75 papers), which mainly employanalytic hierarchy process/analytic network process (AHP/ANP)(20 papers) and linear programming/mixed integer linear pro-gramming (LP/MILP) (18 papers) as solution approaches, are themore popular model purposes when compared to descriptivemodels (59 papers). Descriptive modeling efforts are most oftenutilizing systemic models (39 papers), in particular life cycle anal-ysis (LCA) (24 papers), inputoutput-analysis (IOA) (4 papers), andmetrics (8 papers) as solution approaches. Combining the SCM per-spectives and the modeling categories (see Table 9) shows thatmanagerial decision making is often supported by optimizationmethods while in macroscopic contexts, models are most oftenemployed to explore or explain the interplay of various factors.On the intra-organizational level of a single rm or a particularfunction, normative SSCMmodels are more often employed (42 pa-pers) than descriptive models (24 papers). In contrast, descriptivemodels (20 papers) represent the vast majority of 25 papers that

Process of analysis Functional application area

Plan 53 Logistics 9Source 12 Network design 13Make 18 Outsourcing/offshoring 2Deliver 9 Planning 3Return 0 Product dev. 9Various/other 0 Production 40n.a. 42 Sourcing 12

SCM 16Technology 5Various/other 6n.a. 15Singulara 4

considered in one paper each.

perational Research 233 (2014) 299312contain models for industrial sectors or macro-economies.Stochastic approaches (12 normative, 3 descriptive) are not well

represented in this literature. Normative stochastic models mostoften focus on manufacturers and strive for improving company-wide planning processes (Chen & Fan, 2012; Hu & Bibanda, 2009;Linninger, Chakraborty, & Colberg, 2000; Radulescu et al., 2009;Tsai & Hung, 2009; Wu & Chang, 2004) or the implementation ofcleaner production technologies (Tseng, Lin, & Chiu, 2009). Otherapproaches deal with network design at a carrier (Lee et al.,2010) or the product sustainability assessment in the automotiveindustry (Ghadimi, Azadnia, Yusof, & Saman, 2012). Normative sto-chastic models that take an inter-organizational perspective focuson green SCM in the electronics industry networks (Hsu & Hu,2008; Che, 2010) or on sustainability policies in a macroeconomiccontext (Munda, 2009). Descriptive stochastic models are all atthe inter-organizational level (Kainuma & Tawara, 2006;Saint Jean, 2008; van der Vorst, Tromp, & van der Zee, 2009). Thelimited amount of stochastic models provides opportunities forfurther SSCM research. But uncertainty may also be investigated

Modeling techniquea Solution approachb

77 Articial intelligence 5 AHP/ANP 207 DES 1 DEA 31 Game theory 1 Goal programming 2

36 Meta-heuristics 1 IOA 494

, cas, simed in

a Bicycle, biotechnology, health care, mining, packaging, petroleum, and utilitieswere considered in one paper each.

of Operational Research 233 (2014) 299312 307using various scenarios, each evaluated with a deterministic

Table 8Evaluation results structural dimension Modeling.

Model purpose Model type

Descriptive deterministic 56 AnalyticalDescriptive stochastic 3 HeuristicsNormative deterministic 63 HybridNormative stochastic 12 Mathematical programming

SimulationVarious/other

a Not employed in reviewed papers: spreadsheet calculation, n.a.b Not employed in reviewed papers: Ant colony optimization, Bayesian networks

search procedure, neural networks, petri net, particle swarm optimization, queuingc Dynamic programming, genetic algorithms, and neural networks were consider

Table 9Evaluation results combination of categories Level of analysis and Modelpurpose.

Level of analysis Normative models Descriptive models Total

Function, rm 42 24 66Dyad, chain, network 24 14 38Industry, macro-econ. 5 20 25n.a., other 4 1 5

Total 75 59 134

M. Brandenburg et al. / European Journalapproach.Multi-criteria decision making (MCDM) (25 papers) and mathe-

matical programming (35 papers), with only two papers (Gunson,Klein, Veiga, & Dunbar, 2010; Yura 1994) optimizing a single-objective function, are most often chosen model types and tech-niques to investigate SSCM. These results may occur because (1)the application nature of the journals supports more normative ap-proaches; and (2) the multi-dimensionality of the sustainabilityproblem requires integrating a variety of factors and objectivessimultaneously. A hybrid model is developed by Fichtner, Frank,and Rentz (2004), who combine a MILP approach for economicoptimization with a technical process simulation and a LCA forthe ecologic assessment of energy supply systems.

Surprisingly, solution approaches based on genetic algorithms(Radulescu et al., 2009) as well as dynamic programming (Hu &Bibanda, 2009), goal programming (Tsai & Hung, 2009; Yura,1994), and neural networks (Kuo, Wang, & Tien, 2010) models occur only in very few papers. This result is in contrast to Ilginand Guptas (2010) observation of modeling approaches for green(re-)manufacturing and product recovery.

4.2.3. Research dimensionFor the industry focus (see Table 10), model-based SSCM re-

search focuses on technology-related sectors (39 papers on energy,electronics, or automotive), consumable goods (18 papers on agri-cultural or food & beverage industry) or on macro-economic con-texts (10 paper), while surprisingly, from an environmentalperspective, the transportation industry (Lee, Geum, Lee, & Park,2012; Lovric et al., 2012; with a macroscopic view: lengin, Kabak,nsel, lengin, & Aktas, 2010) has not been as well covered in thisliterature. Clearly, there is room in a number of industries for for-mal models to be applied. The imbalance in industrial perspectivescould be due to variations in the relative importance ofsustainability topics or to convenience and data availability. ButMcdm 25 LCA 24Multi-objective 34 LP/MILP 18Simple heuristics 1 Metrics 8Single-objective 2 Nonlinear programming 5System dynamics 3 Rough set 3Systemic models 39 Variation inequality 6Various/other 22 Various/other 31

n.a. 7Singularc 3

e-based reasoning, differential evolution, fuzzy logic, greedy randomized adaptiveulated annealing.one paper each.

Table 10Evaluation results structural dimension Research (number of studies perindustry).

Agriculture 10 Energy 15 Retail 3Apparel 3 Food &

beverages8 Transportation 2

Automotive 8 Furniture 2 Various/other 6Chemical/

pharmaceutical2 Macro-economy 10 n.a. 32

Construction 2 Metal 6 Singulara 7Electronics 16 Paper 2the diffusion of modeling efforts to other industries is clearly avaluable research direction.

In SSCMmodeling papers, the numerical analysis (see Table 11)is based on empirical data (105 papers), but most of these numer-ical sections may not be easily replicable and are with limited rigorin the underlying empirical research methods. This result is in linewith Seurings (2012) observations and points towards the need tocombine empirical and model-based research as addressed by Has-sini et al. (2012).

Furthermore, specic areas for generalizable future researchare seldom addressed (only 24 papers, see Table 11). These deduc-tively identied research directions, which include recommenda-tions on research design, models and metrics, resources andprocesses, and interfaces with customers and various systems,are now briey summarized. For research design a higher degreeof integration between empirical and theoretical research is rec-ommended (Ukidwe and Bakshi, 2005). A fertile model-based re-search direction is developing new solution methodologies forgreen SC network design (Wang, Lai, & Shi, 2011). Additional mod-eling suggestions include employing asymmetric competitionmodels to assess sustainability triggers (Yalabik & Fairchild,2011) and integrating the decision maker in the evaluation of solu-

Table 11Evaluation results structural dimension Research (data basis and researchperspective).

Data basis of numerical example Research perspective

Empirical data 105 Extend/validate 73None 3 None 37Generic example 26 Specic 24

tions (Harris, Naim, Palmer, Potter, & Mumford, 2011). Tools tomodel ecological factors in processes and the denition and imple-mentation of related effective ecological metrics have also beenrecommended (Smith and Ball, 2012).

As mentioned in Section 2.2, transportation has been neglectedin SSCM modeling. Metrics are needed to incorporate specic fuelconsumption gures for transportation (Harris et al., 2011) or toconsider transportation modes or demand uncertainties (Wanget al., 2011). Addressing opportunity costs to assess environmentalimpacts and their tradeoffs are additional issues that can be incor-porated in various models (Figge & Hahn, 2012; Mouzon, Yildirim,& Twomey, 2007).

For resources and processes, further understanding of resourcereduction impacts on production activities is needed (Smith & Ball,2012). Innovative production process investigation, such as thedevelopment of machines with multiple sleep mode states, prod-uct design process improvements, eco-design technology initia-tives and their evaluation, are all directions (Bovea & Wang,

pers in total). Other papers investigate the design of products(Andersson et al., 1998) or networks (Gunson et al., 2010) or aretechnology-related (Kaldellis, Simotas, Zarakis, & Kondili, 2009;Kiwjaroun, Tubtimdee, & Piumsomboon, 2009; Saint Jean, 2008).More general approaches for environmental management prac-tices are suggested by Sarkis (1998). Most of these 41 manuscriptswere published after 2007 (27 papers), only six papers before 1999and eight between 2002 and 2006. Sixty-one papers from the sam-ple deal with the interface between environmental and economicissues and can be labeled eco-efciency papers, which cover nor-mative (36 papers) and descriptive models (25 papers) in a deter-ministic (51 papers) as well as in a stochastic way (10 papers). Allactors, levels, processes, and most of the functions of analysis areevaluated on eco-efciency. The green research areas are com-prehensive and cover about 90% of all 67 manufacturer-relatedSSCM models.

Compared to the extensive model-based research on environ-mental issues of SSCM, the social aspects are neglected. Only fourpapers elaborate on social issues (Yura, 1994) or the socio-eco-

308 M. Brandenburg et al. / European Journal of Operational Research 233 (2014) 2993122003; Kengpol & Boonkanit, 2011; Mouzon et al., 2007). Assessingand improving the manufacturer-retailer-interface to eliminatesources of environmentally unsustainable practices, and relatingthese to consumer utility and reactions from rms is needed(Darlington & Rahimifard, 2007; Yalabik & Fairchild, 2011; Feng,Li, Duan, & Zhang, 2007). In addition, sustainability benets ofcollaboration, coordination, information sharing, and communicationwithin an SC can be assessed further using quantitative methods (e.g.see Kainuma & Tawara, 2006).

4.2.4. Sustainability dimensionThis dimension focuses primarily on social and environmental

factors of sustainability and their interplay and overlap with eachother and the economic factor. Most formal SSCM models includeenvironmental factors and aspects of eco-efciency while the so-cial dimension is neglected. Holistic models that cover all sustain-ability dimensions have gained attention in the last ten years.Ecological and social dimensions are often modeled using genericfactors, although more specic metrics are employed as well. Themain evaluation results of the sustainability dimension are de-picted in Fig. 3.

Forty-one papers focus on environmental issues exclusivelyand comprise all model purposes, most of the primary actors, allSC levels, and processes of analysis. The functional focus of analysisis predominantly within production (19 papers) or the other tra-ditional SCM functions sourcing, logistics, and SCM (14 pa-Fig. 3. Distribution of papers over sustainability categories (basenomic (Brent, Rogers, Ramabitsa-Siimane, & Rohwer, 2007; Abreu& Camarinha-Matos, 2008) or socio-environmental (Clift, 2003)interfaces.

Twenty-eight papers from the sample set describe holisticSSCM models that cover all three sustainability dimensions. Allof these papers, except Georgopoulou et al. (1998), were publishedafter 2004, and more than half (15 papers) were published withinthe last three years. This indicates that holistic SSCM models rep-resent one of the more recent areas of SSCM research investigation.The large majority of these 28 holistic SSCM models have a norma-tive purpose (21 papers), which is in contrast to the balance be-tween normative and descriptive models observed in all 134reviewed papers (see Section 4.2.2). With regards to the SCM per-spective it is detected that the 28 holistic SSCM models are oftenemployed for macroscopic analyses. Every third holistic SSCMmodel focuses on legal authorities or an industrial sector, whilethese actors of analysis are taken into account in less than 20% ofall 134 reviewed papers (see Section 4.2.1). Furthermore, the highshare of models for manufacturers in the overall paper sample(every second paper, see Section 4.2.1) is not observed in the 28holistic SSCM models; only 7 papers deal with manufacturingrms. These ndings are explained by the circumstance that socialeffects, e.g. employment rate changes or other societal impacts, areusually reected in macroscopic contexts but not on the microeco-nomic level of a rm or function.d on Carter & Rogers, 2008; Kannegiesser & Gnther, 2013).

To conclude the assessment of the sustainability dimension,metrics that were used to represent sustainability factors in for-mal SSCMmodels are summarized. Although sustainability aspectsare often modeled using generic factors such as sustainability cri-teria, social benet, or ecologic impact, more specic metricscan be found for each sustainability dimension. Economic aspectscan be categorized into microeconomic factors such as cost, prot-ability, or revenue (Lovric et al., 2012) and macroeconomic metricsincluding gross domestic product or growth rate (Agrell et al.,2004) as well as labor productivity, market concentration, or im-port dependency (Yakovleva, Sarkis, & Sloan, 2011) or overall

In many of the research approaches evaluated here, research

surveys are difcult to complete for multi-tier supply chains, mod-

M. Brandenburg et al. / European Journal of Op4.3. A research model and additional considerations

By aggregating the various research evaluation dimensions dis-cussed, a research model can be formed by integrating the corearguments of the analysis together (see Fig. 4). The ultimatechoice/solution begins with consideration of the SCM-dimensionsmodeled, which are inuenced by the respective industrial context.The sustainability aspects to be considered then serve as a kind ofmoderating variable that drives the modeling purpose. This ap-proach is based on Seurings (2012) contention that trade-offsamong the economic and environmental dimension are keyassumptions for much of the model building. These tradeoffs canoccur amongst any combination of the sustainability dimensions.The high share of multi-criteria decision making models underpinsthis contention, as the trade-offs serve as a starting point that iseasier to model. Considering the broader set of tradeoffs expandsthe solution space. Many researchers have argued for sustainabilityevaluation in sustainable supply chains among the economic andsocial dimension (or) the environmental and social dimensionmacro-economic development (Feng, Li, Duan, & Zhang, 2007).Environmental aspects comprise input-oriented factors includingrenewable energy sources (Georgopoulou, Saradis, & Diakoulaki,1998; Munda, 2009), natural resources (Liu, Li, Wang, & Dong,2011), water and energy consumption (Yakovleva et al., 2011), orwater quality (Feng et al., 2007). Output-oriented environmentalfactors focus on waste (Yakovleva et al., 2011) and pollution (Geor-gopoulou et al., 1998). Beyond this, ecological factors can be distin-guished into environmental impacts of construction, normaloperations, and failure (Dey, 2006). Specic social aspects are re-lated to internal factors such as wages, employees, or employmentgender ratios (Yakovleva et al., 2011) and furthermore point to-wards external inuences including individual customer needsand requirements (Lovric et al., 2012), social acceptance andcontribution to employment (Georgopoulou et al., 1998), orpopulation growth (Feng et al., 2007). There are papers such asHandeld et al. (2002) or Hassini et al. (2012) that suggest a broadset of environmental factors which can be grouped at multiplelevels.Fig. 4. An aggregated research model.eling efforts are more exible in the number of players involvedand allow evaluating sustainable supply chains. As the eld contin-ues to mature, more complex and insightful modeling can beintegrated.

Expanding the development of criteria sets for sustainable sup-plier selection to integrate environmental and especially social as-pects is required. The challenge is a sufciently comprehensive andprecise simultaneous modeling effort, such that the solutions arenot trivial, but still solvable problems.

From an industry modeling perspective, the lack of specicindustry focused studies on sensitive industries is especially sur-prising. For example, the transportation industry with its heavycarbon emissions, energy, and materials usage is relatively sparselyrepresented. The chemical/pharmaceutical industry with its poten-tially hazardous waste management is an environmental dimen-sion that seems to be overlooked. Finally, the apparel and textileindustry with its prevalence of social issues (e.g. underpaid work-ers, unsafe and dangerous working conditions) would be a primeconsideration for social sustainability issues. How to integrateand develop models into each of these areas requires careful con-sideration of the intangibility of the measures and modeling.

5. Results and discussion

This paper employs a systematic and methodologically rigorousprocess to review quantitative SSCM models. This study uses con-tent analysis to assess a large sample of related papers and to iden-tify current gaps and future perspectives of model-based SSCMresearch.

Before turning to them, we discuss the contribution of this pa-per. As the title clearly states, the paper describes the body of lit-erature on quantitative models for sustainable supply chainmanagement is which is the core contribution. We show theincreasing publication output and use a number of categories forproviding insights into this body of literature. A general meta-research model is presented as well. This research model allowsfor consideration of how the various review elements t togetherand can aid in further development of a research agenda.

5.1. Research gaps and future research perspectivestended to focus on the production processes of a manufacturingcompany and then analyze the results on a specic function or atthe factory level, with a focus on environmental issues (in line withthe ndings of Seuring, 2012). In these situations why such re-search is published under the label sustainable supply chain man-agement, can be called into question. A more critical perspectivemight be needed avoiding that the conceptual borders of sustain-able SCM are increasingly blurred. Just because a research paperhas utilized and argued that sustainable SCM is being considered,the validity of such suppositions needs to be carefully evaluated.This type of validation will require that a clearer denition of sus-tainable SCM be derived and agreed upon by the researchcommunity.

As a further issue, while there are a number of papers dealingwith supplier selection criteria (e.g. Kuo et al., 2010; Saint Jean,2008), suppliers and the extended supply chain still require con-siderably more attention in respective research. Whereas empirical(e.g. Seuring & Mller, 2008b) and calling for more research onsuch interactions.

erational Research 233 (2014) 299312 309Additional avenues for future research on SSCM models werenot only determined deductively from the research opportunitiessuggested in the assessed papers (see Section 4.2.3) but also induc-

This paper reviewed almost two decades worth of research

of Otively from the results and the discussion of the analysis presentedhere. These inductively identied research directions are related toboth research content and research methodology.

Model-based SSCM research can advance the inter-organiza-tional perspective of SCM and extend this to the level of industrysectors (e.g. Kannegiesser & Gnther, 2013; Kannegiesser, Gnther,& Gylfason, 2013). Economic contributions of vertical coordinationin the SSCM context can be assessed quantitatively. Furthermore,quantitative models could be employed to elaborate on the inter-play of regulatory decisions made by legal authorities and manage-rial decision making in rms, supply chains, or industries.

Beyond the inter-organizational aspects, research gaps can beidentied at the functional SCM level. Taking into account thatmost papers focus on production or general SCM, a thorough anal-ysis of sustainability in transportation and warehousing is recom-mended. Decision making in intermodal transportation to reducegreenhouse gas emissions or vehicle routing under considerationof workers time preferences might be adequate examples for re-lated research questions.

Environmental risk management has been widely neglectedand hence offers signicant potential in model-based SSCM re-search. The lack of social aspects in SSCM models points to-wards future research perspectives in related research, becausesuch factors need to be incorporated in SSCM models. With re-gards to input factors and resource consumption, models thatinclude the option of reducing supply offer another researchperspective.

Further SSCM research is needed with regards to the integra-tion of model-based methods with empirical research, which hasa strong focus on eco-efciency and environmental aspects aswell. Empirical research methods should be employed to identifysocial factors in SCM and their economic prerequisites and impli-cations, while model-based research methods are needed toquantitatively investigate this context. A lack of social sustain-ability research may cause a mistaken impression that holisticsustainability and the TBL concept are simply theoretical con-structs with limited relevance. Employing case study researchrigorously (e.g. Seuring, 2008; Stuart, McCutcheon, Handeld,MacLachlin, & Samson, 2002) is highly recommended not onlyfor SSCM models in particular but also for model-based researchin general. Linking empirical and model-based research, as sug-gested by Golicic, Davis, and McCarthy (2005), could broadenthe scientic eld of SSCM with regards to the focused indus-tries. The food, apparel, or automotive industries represent sec-tors that are thoroughly investigated by only one of these twomethodological designs while being neglected by the other.Expanding this industry focus of the employed scientic ap-proaches would balance this existing research bias. Furthermore,the chemicals & oil sector seems to be neglected so far by bothempirical and model-based research.

Focusing on developments and directions of SSCM models leadsto the question of why some of the sophisticated modeling ap-proaches have been widely neglected so far. Dynamic program-ming, evolutionary algorithms, or local search methods representnormative approaches to solve complex problems that offer largeoptimization potential. The fact that these solution approachesare employed only seldom in context to SSCM models outlines fur-ther research opportunities and raises the question of whether thecomplexity of SSCM problems and their optimization potential arealready fully exposed.

5.2. A comparative analysis to other SSCM literature reviews

310 M. Brandenburg et al. / European JournalThe need for further research on OR applications and hybridqualitative and quantitative approaches (Min & Kim, 2012) as wellas more empirical rigor in numerical analyses of SSCM modelsfocusing on SSCM quantitative, formal modeling. One hundredthirty-four articles were identied and utilized in the analysis ofthe research. Overall, the number of publications in this topicalarea is not as large as empirical and conceptual work, but it isgrowing. The ndings included variations in focus by function, re-search perspective, methodology, and the type of sustainability fo-cus of the supply chains. The results also provided opportunities toidentify gaps in the research that could be addressed and poten-tials for further research directions. Some major ndings for fertileareas of research include the integration of social issues into mod-eling, expanding the scope and diffusing modeling from one indus-try to another, and the need for more stochastic approaches inmodeling to relay a more realistic uncertain decision environmentassociated with these many and complex environmental factorsidentied with SSCM.

Although this study was rigorously completed, there are stilllimitations that we encountered, but these limitations provideopportunity for future research. Despite the fact that severalresearchers were involved in the validation and the content analy-sis of this studys paper sample, the categorization of these papersremains interpretative and hence subjective. Furthermore, statisti-cal methods (e.g. Wolf, 2008) could be employed to cluster the pa-per sample and to analyze contingencies of different categories.Additionally, more comprehensive bibliometric citation analysesrepresent another rigorous and structured approach to assess re-lated scientic literature (e.g. Linnenluecke & Grifths, 2012).These limitations leave room for future analyses and reviews ofSSCM modeling publications.

As can be seen, the SSCM modeling eld is on the research up-(Seuring, 2012) are conclusions reinforced by this study. This studyalso conrms the relative lack of social factors in (model-based)SSCM research (Seuring, 2012; Seuring & Mller, 2008a; Tang &Zhou, 2012). There are some contradictions to other reviews thatnd a strong environmental focus but an increased considerationof social factors in empirical SSCM research (Carter & Easton,2011). We found that the green focus of model-based SSCM re-search remains unchanged.

In further contrast to empirical research, which has predomi-nantly investigated the food, apparel, and consumer products sec-tors (Carter & Easton, 2011; Gold et al., 2010a), the results of thispaper indicate that model-based SSCM research mainly focuseson technology-related industries. This industry focus does supportother SSCM modeling literature reviews (Hassini et al., 2012). Un-like empirical research, which focuses on intra-organizational level(Carter & Easton, 2011), formal SSCM models integrate inter-orga-nizational interdependencies. Beyond this, SSCM models do in-clude the perspective of legal authorities and hence are capableof reecting external triggers of sustainability (Gold et al., 2010a,2010b; Seuring & Mller, 2008a).

In terms of modeling results, the ndings do conrm the highrelevance of AHP/ANP, LCA, and MCDM for SSCM models (Seur-ing, 2012), but DES or GA are not as important for SSCM as forproduct recovery and (re-)manufacturing (Ilgin & Gupta, 2010).The unexploited potential for goal programming in green logis-tics (Dekker et al., 2012) is substantiated for SSCM research bythis study, although the lack of MCDM applications in greenlogistics (Dekker et al., 2012) cannot be conrmed for forwardSSCM.

6. Summary and conclusion

perational Research 233 (2014) 299312swing; signicantly more modeling based research can and needsto be completed to more fully understand and integrate SSCM intobusiness thought and practice.

of OpReferences

Abreu, A., & Camarinha-Matos, L. M. (2008). On the role of value systems to promotethe sustainability of collaborative environments. International Journal ofProduction Research, 46(5), 12071229.

Agrell, P. J., Stam, A., & Fischer, G. W. (2004). Interactive multi-objective agro-ecological land use planning: The Bungoma region in Kenya. European Journal ofOperational Research, 158, 194217.

Athanasopoulou, P. (2009). Relationship quality: A critical literature review andresearch agenda. European Journal of Marketing, 43(5/6), 583610.

Andersson, K., Hogaas Eide, M., Lundqvist, U., & Mattsson, B. (1998). The feasibilityof including sustainability in LCA for product development. Journal of CleanerProduction, 6, 289298.

Bertrand, J. W. M., & Fransoo, J. C. (2002). Operations management researchmethodologies using quantitative modeling. International Journal of Operations& Production Management, 22(2), 241264.

Bettley, A., & Burnley, S. (2008). Towards sustainable operations management:Integrating sustainability management into operations management strategiesand practices. In K. B. Misra (Ed.), Handbook of performability engineering(pp. 875904). London: Springer.

Bloemhof-Ruwaard, J. M., van Beek, P., Hordijk, L., & van Wassenhove, L. N. (1995).Interactions between operational research and environmental management.European Journal of Operational Research, 85(2), 229243.

Bovea, M. D., &Wang, B. (2003). Identifying environmental improvement options bycombining life cycle assessment and fuzzy set theory. International Journal ofProduction Research, 41(3), 593609.

Brent, A. C., Rogers, D. E. C., Ramabitsa-Siimane, T. S. M., & Rohwer, M. B. (2007).Application of the analytical hierarchy process to establish health care wastemanagement systems that minimise infection risks in developing countries.European Journal of Operational Research, 181, 403424.

Carter, C. R., & Easton, P. L. (2011). Sustainable supply chain management: Evolutionand future directions. International Journal of Physical Distribution & LogisticsManagement, 41(1), 4662.

Carter, C. R., & Rogers, D. S. (2008). A framework of sustainable supply chainmanagement: Moving toward new theory. International Journal of PhysicalDistribution & Logistics Management, 38(5), 360387.

Che, Y. H. (2010). Using fuzzy analytic hierarchy process and particle swarmoptimisation for balanced and defective supply chain problems consideringWEEE/RoHS directives. International Journal of Production Research, 48(11),33553381.

Chen, C. W., & Fan, Y. (2012). Bioethanol supply chain system planning under supplyand demand uncertainties. Transportation Research Part E, 48, 150164.

Clift, R. (2003). Metrics for supply chain sustainability. Clean Technologies andEnvironmental Policy, 5, 240247.

Cooper, M. C., Lambert, D. M., & Pagh, J. D. (1997). Supply chain management Morethan a new name for logistics. International Journal of Logistics Management,8(1), 114.

Daniel, S. E., Diakoulaki, D. C., & Pappis, C. P. (1997). Operations research andenvironmental planning. European Journal of Operational Research, 102(2),248263.

Darlington, R., & Rahimifard, S. (2007). Hybrid two-stage planning for food industryoverproduction waste minimization. International Journal of ProductionResearch, 45(1819), 42734288.

Dekker, R., Bloemhof, J., & Mallidis, J. (2012). Operations Research for green logistics An overview of aspects, issues, contributions and challenges. European Journalof Operational Research, 219, 671679.

Dey, P. K. (2006). Integrated project evaluation and selection using multiple-attribute decision-making technique. International Journal of ProductionEconomics, 103, 90103.

Drake, D., & Spinler, S. (2013). Sustainable operations management: An enduringstream, or passing fancy? Working paper 13-084. Harvard Business School.

Edwards, J. B., McKinnon, A. C., & Cullinane, S. L. (2010). Comparative analysis of thecarbon footprints of conventional and online retailing A last mileperspective. International Journal of Physical Distribution & LogisticsManagement, 40(1/2), 103123.

Elkington, J. (1998). Cannibals with forks: The triple bottom line of the 21st century.Stoney Creek/CT: New Society.

Elkington, J. (2004). Enter the triple bottom line. In A. Henriques & J. Richardson(Eds.), The triple bottom line: Does it all add up? (pp 116). London: Earthscan.

Feng, S., Li, L. X., Duan, Z. G., & Zhang, J. L. (2007). Assessing the impacts of South-to-North water transfer project with decision support systems. Decision SupportSystems, 42, 19892003.

Fichtner, W., Frank, M., & Rentz, O. (2004). Inter-rm energy supply concepts: Anoption for cleaner energy production. Journal of Cleaner Production, 12, 891899.

Figge, F., & Hahn, T. (2012). Is green and protable sustainable? Assessing the trade-off between economic and environmental aspects. International Journal ofProduction Economics. http://dx.doi.org/10.1016/j.ijpe.2012.02.001.

Fleischmann, M., Bloemhof-Ruwaard, J. M., Dekker, R., van der Laan, E., van Nunen, J.A. E. E., & van Wassenhove, L. N. (1997). Quantitative models for reverselogistics: A review. European Journal of Operational Research, 103(1), 117.

Gareld, E. (2004). Historiographic mapping of knowledge domains literature.Journal of Information Science, 30, 119145.

M. Brandenburg et al. / European JournalGeorgopoulou, E., Saradis, Y., & Diakoulaki, D. (1998). Design and implementationof a group DSS for sustaining renewable energies exploitation. European Journalof Operational Research, 109, 483500.Ghadimi, P., Azadnia, A. H., Yusof, M. N., & Saman, M. Z. M. (2012). A weighted fuzzyapproach for product sustainability assessment: A case study in automotiveindustry. Journal of Cleaner Production, 33, 1021.

Gold, S., Seuring, S., & Beske, P. (2010a). The constructs of sustainable supply chainmanagement A content analysis based on published case studies. Progress inIndustrial Ecology An International Journal, 7(2), 114137.

Gold, S., Seuring, S., & Beske, P. (2010b). Sustainable supply chain management andinter-organizational resources: A literature review. Corporate SocialResponsibility and Environmental Management, 17, 230245.

Golicic, S. L., Davis, D. F., & McCarthy, T. M. (2005). A balanced approach to researchin supply chain management. In H. Kotzab, S. Seuring, M. Mller, & G. Reiner(Eds.), Research methodologies in supply chain management (pp. 89108).Heidelberg: Physica.

Golicic, S. L., & Smith, C. D. (2013). A meta-analysis of environmentally sustainablesupply chain management practices and rm performance. Journal of SupplyChain Management, 49(2), 7895.

Guide, V. D. R., & van Wassenhove, L. N. (2009). The evolution of closed-loop supplychain research. Operations Research, 57(1), 1018.

Gungor, A., & Gupta, S. M. (1999). Issues in environmentally consciousmanufacturing and product recovery: A survey. Computers & IndustrialEngineering, 36(4), 811853.

Gunson, A. J., Klein, B., Veiga, M., & Dunbar, S. (2010). Reducing mine water networkenergy requirements. Journal of Cleaner Production, 18, 13281338.

Halldorsson, A., & Arlbjrn, J. (2005). Research methodologies in supply chainmanagement What do we know? In H. Kotzab, S. Seuring, M. Mller, & G.Reiner (Eds.), Research methodologies in supply chain management (pp. 107122).Heidelberg: Physica.

Handeld, R.,Walton, S. V., Sroufe, R., &Melnyk, S. A. (2002). Applying environmentalcriteria to supplier assessment: A study in the application of the analyticalhierarchy process. European Journal of Operational Research, 141, 7087.

Harris, I., Naim, M., Palmer, A., Potter, A., & Mumford, C. (2011). Assessing theimpact of cost optimization based on infrastructure modeling on CO2 emissions.International Journal of Production Economics, 131, 313321.

Hassini, E., Surti, C., & Searcy, C. (2012). A literature review and a case study ofsustainable supply chain management with a focus on metrics. InternationalJournal of Production Economics. http://dx.doi.org/10.1016/j.ijpe.2012.01.042.

Hsu, C. W., & Hu, A. H. (2008). Green supply chain management in the electronicindustry. International Journal of Environmental Science and Technology, 5(2),205216.

Hu, G., & Bibanda, B. (2009). Modeling sustainable product life cycle decisionsupport systems. International Journal of Production Economics, 122, 366375.

Ilgin, M. A., & Gupta, S. M. (2010). Environmentally conscious manufacturing andproduct recovery (ECMPRO). A review of the state of the art. Journal ofEnvironmental Management, 91, 563591.

Kainuma, Y., & Tawara, N. (2006). A multiple attribute utility theory approach tolean and green supply chain management. International Journal of ProductionEconomics, 101, 99108.

Kaldellis, J. K., Simotas, M., Zarakis, D., & Kondili, E. (2009). Optimum autonomousphotovoltaic solution for the Greek islands on the basis of energy pay-backanalysis. Journal of Cleaner Production, 17, 13111323.

Kannegiesser, M., & Gnther, H.-O. (2013). Sustainable development of globalsupply chainsPart 1: Sustainability optimization framework. Flexible Servicesand Manufacturing Journal. http://dx.doi.org/10.1007/s10696-013-9176-5.

Kannegiesser, M., Gnther, H.-O., & Gylfason, O. (2013). Sustainable development ofglobal supply chainsPart 2: Investigation of the European automotiveindustry. Flexible Services and Manufacturing Journal. http://dx.doi.org/10.1007/s10696-013-9177-4.

Kengpol, A., & Boonkanit, P. (2011). The decision support framework for developingEcodesign at conceptual phase based upon ISO/TR 14062. International Journal ofProduction Economics, 131, 414.

Kiwjaroun, C., Tubtimdee, C., & Piumsomboon, P. (2009). LCA studies comparingbiodiesel synthesized by conventional and supercritical methanol methods.Journal of Cleaner Production, 17, 143153.

Kleijnen, P. (2005). Supply chain simulation tools and techniques: A survey.International Journal of Simulation and Process Modelling, 1(1/2), 8289.

Kleindorfer, P. A., Singhal, K., & van Wassenhove, L. N. (2005). Sustainableoperations management. Production and Operations Management, 14(4),482492.

Krippendorff, K. (1980). Content analysis. Beverly Hills/CA: Sage.Kuo, R. J., Wang, Y. C., & Tien, F. C. (2010). Integration of articial neural network

and MADA methods for green supplier selection. Journal of Cleaner Production,18, 11611170.

Lage Junior, M., & Godinho Filho, M. (2010). Variations of the Kanban system:Literature review and classication. International Journal of ProductionEconomics, 125, 1321.

Lebreton, B. (2007). Strategic closed-loop supply chain management. Lecture notes ineconomics and mathematical systems. Berlin: Springer. 586.

Lee, D. H., Dong, M., & Bian, W. (2010). The design of sustainable logistics networkunder uncertainty. International Journal of Production Economics, 128, 159166.

Lee, S., Geum, Y., Lee, H., & Park, Y. (2012). Dynamic and multidimensionalmeasurement of product-service system (PSS) sustainability: A triple bottomline (TBL)-based system dynamics approach. Journal of Cleaner Production, 32,173182.

erational Research 233 (2014) 299312 311Linnenluecke, M. K., & Grifths, A. (2012). Firms and sustainability: Mapping theintellectual origins and structures of the corporate sustainability eld. GlobalEnvironmental Change. http://dx.doi.org/10.1016/j.gloenvcha.2012.07.007.

Linninger, A. A., Chakraborty, A., & Colberg, R. D. (2000). Planning of waste reductionstrategies under uncertainty. Computers & Chemical Engineering, 24(27),10431048.

Liu, D., Li, H., Wang, W., & Dong, Y. (2011). Constructivism scenario evolutionaryanalysis of zero emission regional planning: A case of Qaidam Circular EconomyPilot Area in China. International Journal of Production Economics. http://dx.doi.org/10.1016/j.ijpe.2011.04.008.

Lovric, M., Li, T., & Vervest, P. (2012). Sustainable revenue management: A smartcard enabled agent-based modeling approach. Decision Support Systems. http://dx.doi.org/10.1016/j.dss.2012.05.061.

Mayring, P. (2002). Qualitative Sozialforschung (Qualitative social research) (5th ed.).Weinheim: Beltz.

Mayring, P. (2008). Qualitative Inhaltsanalyse (Qualitative content analysis) (10th ed.).Weinheim: Beltz.

Mentzer, J., DeWitt, W., Keebler, J., Min, S., Nix, N., Smith, C., & Zacharia, Z. (2001).Dening supply chain management. Journal of Business Logistics, 22(2),

Seuring, S., & Mller, M. (2008a). From a literature review to a conceptualframework for sustainable supply chain management. Journal of CleanerProduction, 16(15), 16991710.

Seuring, S., & Mller, M. (2008b). Core issues in sustainable supply chainmanagement A Delphi study. Business Strategy and the Environment, 17(8),455466.

Shapiro, J. (2007). Modeling the supply chain (2nd ed.). Duxbury: Brooks/Cole,Thompson.

Sheu, J. B., Chou, Y. H., & Hu, C. C. (2005). An integrated logistics operational modelfor green-supply chain management. Transportation Research Part E, 41,287313.

Smith, L., & Ball, P. (2012). Steps towards sustainable manufacturing throughmodelling material, energy and waste ows. International Journal of ProductionEconomics. http://dx.doi.org/10.1016/j.ijpe.2012.01.036.

Srivastava, S. K. (2007). Green supply-chain management: A state-of-the-art

312 M. Brandenburg et al. / European Journal of Operational Research 233 (2014) 299312125.Meredith, J. (1993). Theory building through conceptual methods. International

Journal of Operations & Production Management, 13(3), 311.Mikkola, J. H. (2005). Modeling the effect of product architecture modularity in

supply chains. In H. Kotzab, S. Seuring, M. Mller, & G. Reiner (Eds.), Researchmethodologies in supply chain management (pp. 493508). Heidelberg: Physica.

Min, H., & Kim, I. (2012). Green supply chain research: Past, present, and future.Logistics Research, 4, 3947.

Mouzon, G., Yildirim, M. B., & Twomey, J. (2007). Operational methods forminimization of energy consumption of manufacturing equipment.International Journal of Production Research, 45(1819), 42474271.

Munda, G. (2009). A conict analysis approach for illuminating distributional issuesin sustainability policy. European Journal of Operational Research, 194, 307322.

Radulescu, M., Radulescu, S., & Radulescu, C. Z. (2009). Sustainable productiontechnologies which take into account environmental constraints. EuropeanJournal of Operational Research, 193, 730740.

ReVelle, C. (2000). Research challenges in environmental management. EuropeanJournal of Operational Research, 121, 218231.

Saaty, T. L. (1980). The analytic hierarchy process: Planning, priority setting, resourceallocation. Pittsburgh/PA: RWS Publications.

Saint Jean, M. (2008). Polluting emissions standards and clean technologytrajectories under competitive selection and supply chain pressure. Journal ofCleaner Production, 16S1, S113S123.

Sarkis, J. (1998). Evaluating environmentally conscious business practices. EuropeanJournal of Operational Research, 10, 159174.

Sarkis, J. (2012). A boundaries and ows perspective of green supply chainmanagement. Supply ChainManagement: An International Journal, 17(2), 202216.

Sarkis, J., Zhu, Q., & Lai, K. H. (2011). An organizational theoretic review of greensupply chain management literature. International Journal of ProductionEconomics, 130(1), 115.

Sasikumar, P., & Kannan, G. (2008a). Issues in reverse supply chains, part I: End-of-life product recovery and inventory management An overview. InternationalJournal of Sustainable Engineering, 1(3), 154172.

Sasikumar, P., & Kannan, G. (2008b). Issues in reverse supply chain, part II: Reversedistribution issues An overview. International Journal of SustainableEngineering, 1(4), 234249.

Sasikumar, P., & Kannan, G. (2009). Issues in reverse supply chain, part III:Classication and simple analysis. International Journal of SustainableEngineering, 2(1), 227.

Sbihi, A., & Eglese, R. W. (2007). Combinatorial optimization and green logistics.4OR, 5, 99116.

Seuring, S. (2008). Assessing the rigor of case study research in supply chainmanagement. Supply Chain Management: An International Journal, 13(2),128137.

Seuring, S. (2012). A review of modeling approaches for sustainable supply chainmanagement. Decision Support Systems. http://dx.doi.org/10.1016/j.dss.2012.02.053.

Seuring, S., & Gold, S. (2012). Conducting content-analysis based literature reviewsin supply chain management. Supply Chain Management: An InternationalJournal, 17(5), 544555.literature review. International Journal of Management Reviews, 9(1), 5380.Stock, J. R., & Boyer, S. L. (2009). Developing a consensus denition of supply chain

management: A qualitative study. International Journal of Physical Distribution &Logistics Management, 39(8), 690711.

Stuart, I., McCutcheon, D., Handeld, R., MacLachlin, R., & Samson, D. (2002).Effective case research in operations management: A process perspective.Journal of Operations Management, 20, 419433.

Supply-Chain Council (Ed.), (2008). Supply chain operations reference model SCORversion 9.0. Pittsburgh: Supply-Chain Council.

Tang, C. S., & Zhou, S. (2012). Research advances in environmentally and sociallysustainable operations. European Journal of Operational Research, 223, 585594.

Tsai, W. H., & Hung, S. J. (2009). A fuzzy goal programming approach for greensupply chain optimisation under activity-based costing and performanceevaluation with a value-chain structure. International Journal of ProductionResearch, 47(18), 49915017.

Tseng, M. L., Lin, Y. H., & Chiu, A. S. F. (2009). Fuzzy AHP-based study of cleanerproduction implementation in Taiwan PWB manufacturer. Journal of CleanerProduction, 17, 12491256.

Ukidwe, N. U., & Bakshi, B. R. (2005). Flow of natural versus economic capital inindustrial supply networks and its implications to sustainability. EnvironmentalScience Technology, 39, 97599769.

lengin, F., Kabak, ., nsel, S., lengin, B., & Aktas, E. (2010). A problem-structuring model for analyzing transportationenvironment relationships.European Journal of Operational Research, 200, 844859.

van der Vorst, J. G. A. J., Tromp, S. O., & van der Zee, D.