The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0960-0035.htm

IJPDLM
41,1 Sustainable supply chain
management: evolution and
future directions
46
Craig R. Carter
Arizona State University, Tempe, Arizona, USA, and
P. Liane Easton
Center for Logistics Management, University of Nevada, Reno, Nevada, USA

Abstract
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

Purpose – The purpose of this paper is to conduct a systematic review of the sustainable supply
chain management (SSCM) literature in the principal logistics and supply chain management journals,
across a 20-year time frame.
Design/methodology/approach – The authors use a systematic literature review methodology.
This methodology allows for the minimization of researcher bias and the maximization of reliability
and replicability. The study’s empirical validity is further enhanced by demonstrating high levels of
inter-coder reliability across families of codes.
Findings – The field of SSCM has evolved from a perspective and investigation of standalone
research in social and environmental areas; through a corporate social responsibility perspective; to
the beginnings of the convergence of perspectives of sustainability as the triple bottom line and the
emergence of SSCM as a theoretical framework. While the SSCM research has become more
theoretically rich and methodologically rigorous, there are numerous opportunities for further
advancing theory, methodology, and the managerial relevance of future inquiries.
Research limitations/implications – The trends and gaps identified through our analysis allow
us to develop a cogent agenda to guide future SSCM research.
Practical implications – The current perspectives of SSCM hold important implications for
managers, by directing limited resources toward projects which intersect environmental and/or social
performance, and economic performance.
Originality/value – The paper provides a systematic, rigorous, and methodologically valid review
of the evolution of empirical SSCM research across a 20-year time period.
Keywords Supply chain management, Economic sustainability, Social responsibility,
Environmental management, Economic performance
Paper type Literature review

Introduction
Sustainability has become a huge buzzword, both in today’s business world and
within the broader facets of society. It is difficult, for example, to walk by a newsstand
without seeing at least one magazine cover featuring alternative sources of energy,
climate change issues, or the iconic polar bear floating on a thin sheet of ice. There are a
number of drivers for this rising prominence of sustainability, including supply and
International Journal of Physical demand characteristics surrounding energy consumption, an increased understanding
Distribution & Logistics Management of the science relating to climate change, and greater transparency concerning both the
Vol. 41 No. 1, 2011
pp. 46-62 environmental and the social actions of organizations.
q Emerald Group Publishing Limited These issues are relevant to managers, because their stakeholders – customers,
0960-0035
DOI 10.1108/09600031111101420 regulatory bodies, non-governmental organizations, and even their own employees – are
 increasingly demanding that organizations address and manage the environmental SSCM
and social issues which are impacted by their operations. Supply chain managers
are in a particularly advantageous position to impact – positively or negatively –
environmental and social performance, through for example supplier selection and
supplier development, modal and carrier selection, vehicle routing, location decisions,
and packaging choices.
The purpose of this paper is to provide a systematic review of the evolution of 47
sustainable supply chain management (SSCM) over the past 20 years, with the goal of
identifying trends, potential consensus in findings and approaches across studies, and
gaps, all of which can help to guide future research and improve the management of
sustainable supply chain initiatives. In addition, we will take this opportunity to share
our own perspectives regarding the current state of the field and where we believe
research should be heading.
The remainder of the paper is organized as follows. In the next section, we describe the
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

evolution of the research in, and management of, SSCM. We then describe the systematic
literature review methodology and our specific application of this methodology.
We present our findings in the final section of the paper, and we discuss the research
opportunities and managerial implications associated with each set of findings.

Summary of the literature

The conceptualization and management of social and environmental issues has
evolved from what we term “standalone,” through the notion of social responsibility,
and finally to the concept of sustainability. We describe this evolution next.
Much of the supply chain research that many would consider to be a part of
corporate social responsibility (CSR) and sustainability has occurred in a standalone
fashion, in which there has been little or no recognition of the interrelationships among
topics such as the environment, diversity, human rights, philanthropy, and safety, and
the fact that these are really components of the larger, more holistic concepts of CSR, and
sustainability as they apply to supply chain management (Carter and Jennings, 2002).
Similarly, supply chain managers often initiated and managed past projects
in a standalone fashion, without a clear, holistic, and more strategic understanding
of how these pieces of the puzzle fit together to create their organization’s overall
sustainability position. Managers also often overlooked opportunities to learn from the
successes and failures of one type of initiative – say in the environmental arena – and
apply this knowledge to future projects in other parts of their organization and in other
areas of sustainability, such as diversity and safety issues (Carter, 2005).
The works of Carter and Jennings (2002) and Murphy and Poist (2002) were among
the earlier papers to view these standalone activities within a broader conceptualization
of CSR. Carter and Jennings (2002) incorporate the social responsibility literature from
the management field and utilize Carroll’s (1979, 1991) hierarchy of economic, legal,
ethical, and discretionary responsibility to place the standalone supply chain
management activities within the context of discretionary activities and thus social
responsibility. Murphy and Poist (2002, p. 23) also position the standalone activities
within the rubric of social responsibility and note the need to “seek socially beneficial
results along with economically beneficial ones.” Both of these papers placed
environmental as well as social (e.g. diversity, philanthropy, human rights, and safety)
activities within the rubric of social responsibility.
 IJPDLM Despite the incorporation of economic responsibility by these and other
researchers who adopt a social responsibility paradigm, much of the extant literature
41,1 in the area of “logistics social responsibility” has operationalized the environmental
and/or social dimensions of CSR, without explicitly accounting for economic
performance. Within industry, managers have often viewed social responsibility as
just that – a responsibility – which did not necessarily yield financial rewards (Walley
48 and Whitehead, 1994). While it is somewhat speculative on our part, possible reasons for
these misperceptions include the lack of a more explicit incorporation of economic
performance into social responsibility frameworks, lack of transparency (e.g. managing
corporations during the pre-internet or early-internet time periods), and/or simply the
use of the term “responsibility” and how this was presented to and perceived by
managers and academics alike.
Carter and Rogers (2008) use a conceptual theory building approach to synthesize
the literature at the time and incorporate complementary theoretical bases to introduce
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

a theoretical framework of sustainability as it is applied to the supply chain – a

concept which they refer to as SSCM (Figure 1). At the core of this conceptualization
is Elkington’s (1998) triple bottom line: the intersection of environmental, social,
and economic performance. This is a parsimonious and potentially powerful means
of conveying what sustainability means for an organization. Rather than suggesting
that firms identify and engage in social and environmental activities which will
hopefully help, or at least not harm, economic performance, the triple bottom line
explicitly directs managers to identify those activities which improve economic
performance and dictate the avoidance of social and environmental activities which fall

Strategy Good? Organizational

culture
• Sustainability as part • Deeply ingrained
of an integrated • Organizational
strategy citizenship
Environmental Social • Values and ethics
performance performance

Sustainability

Best

Better Better

Risk management Transparency

Figure 1.
Sustainable supply chain • Contingency planning Economic • Stakeholder engagement
management • Supply disruptions performance • Supplier operations
• Outbound supply chains
 outside of this intersection. Carter and Rogers (2008) suggest that engaging in SSCM
sustainability, and SSCM in particular, is not discretionary, but rather a requirement.
SSCM involves the long-run improvement of an organization’s economic bottom line
and helps managers to answer the question of, “What is it that we need to do, not just
to survive, but to thrive, and not just one year, three years, or five years from now, but
in ten years, 20 years, and beyond?” Again, this is a salient conceptualization that can
begin to allow managers to take tangible actions. 49
Examples of the activities that fall within the triple bottom line include cost savings
associated with reduced packaging and more effective design for reuse and recycling;
lower health and safety costs, as well as reduced turnover and recruitment costs due
to safer warehousing and transport and improved working conditions; reduced labor
costs in the form of higher levels of motivation and productivity and less absenteeism
resulting from improved working conditions; lower costs, shorter lead-times, improved
product quality, and lower disposal costs resulting from the implementation of ISO
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

14000 standards and the use of design for disassembly and reuse; and an enhanced
organizational reputation, which can make a firm more attractive to both customers
and suppliers (Carter and Rogers, 2008, pp. 370-1).
Carter and Rogers identify four supporting facets, or facilitators of SSCM, which are
also shown in Figure 1:
(1) strategy – holistically and purposefully identifying individual SSCM initiatives
which align with and support the organization’s overall sustainability strategy;
(2) risk management, including contingency planning for both the upstream and
the downstream supply chain;
(3) an organizational culture which is deeply ingrained and encompasses
organizational citizenship, and which includes high ethical standards and
expectations (a building block for SSCM) along with a respect for society (both
within and outside of the organization) and the natural environment; and
(4) transparency in terms of proactively engaging and communicating with key
stakeholders and having traceability and visibility into upstream and
downstream supply chain operations.

We use this conceptualization of sustainability, which Carter and Rogers (2008, p. 368)
define as, “the strategic, transparent integration and achievement of an
organization’s social, environmental, and economic goals in the systemic coordination
of key interorganizational business processes for improving the long-term economic
performance of the individual company and its supply chains,” in the systematic literature
review which we describe next.

Methodology
We used a systematic review methodology to create a dataset of articles and to analyze
those articles. This approach was initially employed largely by medical researchers
and has more recently been adopted in the field of management (Tranfield et al., 2003).
In contrast to a traditional narrative literature review, a systematic review of the
literature reduces researcher bias concerning the inclusion or exclusion of studies, and
clearly communicates how the review was performed (Denyer and Neely, 2004).
 IJPDLM In particular, a systematic review allows for a high level of transparency, with the goal
41,1 of maximizing replicability (Crowther and Cook, 2007).
To narrow the scope to a manageable number of articles, the following seven
logistics and supply chain management journals were selected:
(1) International Journal of Logistics Management.
50 (2) International Journal of Physical Distribution & Logistics Management.
(3) Journal of Business Logistics.
(4) Journal of Operations Management.
(5) Journal of Supply Chain Management.
(6) Transportation Journal.
(7) Transportation Research Part E.
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

These journals are commonly recognized as the top-tier and primary outlets within the
broad field of empirical supply chain management research (Carter et al., 2009;
Giunipero et al., 2008) and they are often incorporated in similar systematic reviews
(Cantor, 2008).
In further narrowing the scope of the literature, the following subjects were
excluded from the review: supply chain security; technical issues surrounding life cycle
analysis, end-of-life, cost modeling, hazardous materials, etc. including the regulatory
issues surrounding these subjects; macro/policy issues (as compared to micro,
company, and supply chain-specific issues); reverse logistics and waste disposal;
consumer issues (e.g. automobile safety); and papers where sustainability was a very
ancillary part of the article’s focus.
In addition, only papers that were empirical in nature were included in the review.
Here, empirical approaches included the collection and analysis of primary or secondary
data and included the use of surveys, case studies, interviews, and laboratory
experiments, as well as conceptual theory building and systematic literature reviews.
Thus, papers that used non-empirical approaches such as mathematical modeling,
and manuscripts which expressed opinions but which relied on anecdotal evidence,
were excluded from the review.
Next, we conducted a manual review of all articles published in these seven
journals, from the period of 1991 through the most current issue that was available as
of June 2010. This period of approximately 20 years was chosen, because:
.
the majority of research which has examined standalone issues relating to SSCM
has been published after 1990 (Carter and Rogers, 2008; Seuring and Müller,
2008); and
.
the empirical research which has examined these supply chain issues
from the more holistic perspectives of CSR and sustainability has been
published almost exclusively during this period (Linton et al., 2007).

This manual review of the literature resulted in the identification of 121 articles
for initial inclusion in the literature database and helped in further refining the
keywords used for the second phase of the search process. In this second phase of
the search process, the authors conducted an electronic review of the literature, using
 the keywords listed in Table I. These keywords were also based on the input of four SSCM
industry professionals who participated in a conference call in early June 2010.
We used the EBSCO Business Source Complete database, with the keywords from
Table I in the All Text and Abstract fields and the seven journal titles in the
Publication Name field. The electronic search process resulted in the inclusion of an
additional 11 papers in the initial database, for a total of 130 papers.
Based on the above criteria, we eliminated 50 of these papers, leaving 80 papers for 51
our analysis. The most common reasons for the elimination of these papers were:
.
a focus on macro-economic/public policy issues (Saltzman and Belzer, 2002)
rather than the actual management of the supply chain;
.
a modeling (Sheu, 2008) as opposed to an empirical methodology; or
.
an article which was an editorial (Jayaraman et al., 2007) or was largely
descriptive in nature (Enarsson, 1998).
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

The basic bibliographic data (authors’ names, title of article, journal name, year of
publication, volume number, issue number, and pages) were then entered into a
database in Microsoft Excel. Next, the lead author coded each article, based on the
coding scheme displayed in Table II.
The second author independently check-coded a subsample of the study’s articles.
We calculated reliability based on the proportion of total pairwise agreements between
the coders. The advantages of this method are that it is easy to calculate and to
understand. In addition, capitalization on spurious agreement between coders should
be minimized given the relatively large number of classification categories (Rust and
Cooil, 1994), and accuracy tends be high due to the relatively large sample size
(Kraemer, 1979)[1].
Given the large number of classification categories, the initial inter-coder agreement
rate is comparable to Cronbach’s (1951) coefficient alpha (Perreault and Leigh, 1989).
The overall inter-coder agreement rate across categories ranged from 93.64 to 100.00
percent, with an average of 97.27 percent. These rates far exceed the 0.70 recommended
minimum and suggest a very high level of reliability and thus replicability of our data
coding.
The results of our data coding are summarized in Table III. Rather than conducting
means test after means test (and suffering from high alpha inflation), we instead focus
on the most impactful and pragmatic differences within each of the categories
displayed in Table III. We discuss these results in the next section of the paper, and in

Carbon Human rights Social enterprise

Corporate social ISO 14000/1 Social responsibility
responsibility Labor Sustainable
Diversity Minority business enterprise/minority women Sustainability
business enterprise
Energy Minority Sweatshop
Environment(al) Philanthropy Triple bottom line
Fair trade Resource conservation Women/women owned
Green Safety Working conditions Table I.
Health Social Search terms
 IJPDLM
Coding family Description of codes
41,1
Subject Standalone area(s): the article investigates a specific dimension(s) of
sustainability (e.g. the environment)
CSR: the article investigates CSR (both environmental and social issues), using
framework(s) from the CSR literature
52 Sustainability: the article investigates the triple bottom line and uses the triple
bottom line as a framework for the article’s conceptualization
Note: an article could be coded as both CSR and sustainability, if both literature
bases are used in developing the article’s conceptualization
Inferential Coded as inferential if the authors use inferential statistics to test hypotheses or
propositions that are stated a priori, or if the authors use an inductive approach
(e.g. grounded theory and rigorous analysis of qualitative data) to develop
explicitly stated propositions or similar, explicitly stated relationships among
variables
Coded as descriptive if the authors take a descriptive approach to their analyses
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

(e.g. present summary statistics; compare means; or even use inferential

statistics such as analysis of variance (ANOVA), but without testing
hypotheses or propositions which were developed and presented prior to the
statistical analysis of the data)
Coded as both if the authors take an exploratory approach, by, for example,
presenting a research objective(s)/research question(s), but they do not
explicitly present hypotheses or propositions which they then test. For example,
the authors may identify potential antecedents and use regression analysis to
see which independent variables “stick”
Moderation Coded as moderation if the authors test for moderation/interaction effects
Methodology and Methodology: this is the primary methodology used to collect the study’s data.
analysis Examples include case study, survey, archival data, literature review, depth
interviews, and focus group interviews
Analysis: the approach(es) used to analyze the study’s data. Examples include
confirmatory factor analysis (CFA), exploratory factor analysis, regression,
ANOVA, structural equation modeling, and qualitative data analysis
Validity: coded as yes if the authors address both reliability and other facets of
validity, somewhat if the authors assess reliability but not other facets of
validity, and no if the authors address neither reliability nor validity
Social desirability bias: coded as yes if the authors address this bias, and no if
they do not
Sample size 1: the number of organizations/firms from which data were
collected
Sample size 2: the total number of respondents/informants (this number is equal
to sample size 1 in the case of a single, key informant, or greater than sample
size 1 in the case of multiple informants)
Unit of analysis: generally, one of the following: the individual, the function or
group, the firm (includes a plant or strategic business unit (SBU)), or the supply
chain. Examples of other include projects and published articles as the unit of
analysis
Context Key informant: the functional affiliation of the key informant (e.g. procurement,
production, distribution, and marketing)
Industry: the investigated industry; coded as multiple in the case of multiple
industries
Theoretical lens(es) Assessed whether the authors use any of the following theories as lenses, or even
Table II. rationale in developing their models: transaction cost economics, the resource-
Coding scheme based view, the knowledge-based view, stakeholder theory, and/or other
 Percenta Percenta Percenta
SSCM
(full time period) (2001-2010) (1991-2000)
(%) (%) (%)

Section A: subject
Environment 42.50 35.42 53.13
Diversity 15.00 4.17 31.25
Human rights/quality of life 6.25 4.17 9.38
53
Safety 21.25 27.08 12.50
Philanthropy 0.00 0.00 0.00
CSR 11.25 18.75 0.00
Sustainability 15.00 25.00 0.00
Section B: industry
Automotive 2.78 4.76 0.00
Consumer products 13.89 11.90 16.67
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

Food and beverage 2.78 2.38 3.33

Transportation 20.83 23.81 16.67
Multi-industry 48.61 47.62 50.00
Other 11.11 9.52 13.33
Section C: theoretical lens(es)
Transaction cost economics 6.25 8.33 3.13
Resource-based view 11.25 16.67 3.13
Knowledge-based view/organizational learning 2.50 4.17 0.00
Stakeholder theory 21.25 35.42 0.00
Other 31.25 45.83 9.38
Multiple lenses 21.25 33.33 3.13
None 55.00 33.33 87.50
Section D: validity
Addressed (reliability and multiple, additional facets
of validity) 45.45 64.44 18.75
Partially addressed (reliability, but no additional
facets of validity) 11.69 8.89 15.63
Not addressed 42.86 26.67 65.63
Section E: social desirability bias
Addressed 15.38 25.00 3.45
Not addressed 84.62 75.00 96.55
Section F: unit of analysis
Individual 20.78 17.78 25.00
Function or group 9.09 8.89 9.38
Firm (including plant or SBU) 62.34 60.00 65.63
Supply chain (at least at dyad) 3.90 6.67 0.00
Other 3.90 6.67 0.00
Section G: methodology
Survey 60.00 47.92 78.13
(Multiple) Case study 17.50 22.92 9.38
Archival data 8.75 10.42 6.25
Empirical/systematic literature review 5.00 8.33 0.00
Conceptual theory building 3.75 4.17 3.13
Focus group interviews 2.50 2.08 3.13
Individual interviews 2.50 4.17 0.00 Table III.
(continued) Results
 IJPDLM Percenta Percenta Percenta
41,1 (full time period) (2001-2010) (1991-2000)
(%) (%) (%)

Section H: analysis
Descriptive statistics (summary statistics,
means testing, and rank analysis) 26.53 11.11 54.29
54 Regression analysis 16.33 19.05 11.43
Qualitative data analysis (data coding, matrix
analysis, content analysis) 15.31 17.46 11.43
CFA 12.24 15.87 5.71
Structural equation modeling, path analysis 9.18 12.70 2.86
EFA 8.16 11.11 2.86
Other 5.10 6.35 2.86
ANOVA 4.08 3.17 5.71
Conceptual theory building 3.06 3.17 2.86
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

Section I: moderation
Moderation/interaction effects 10.26 13.04 6.25
Table III. Note: aPercentages in some sections may not add to 100 percent, due to double coding

doing so, we outline what we believe to be the most pertinent areas of SSCM which
require new and/or additional research.

Discussion: evolution and future directions

Subject
Environmental aspects of supply chain management have been the leading focus
of research over the past 20 years (Section A of Table III). The environment is of
course a key component of the triple bottom line and has been in the media spotlight
concerning climate change and rising energy prices. To some degree as well, the
terms “sustainability” and “environment” have been used interchangeably, both
by researchers and by managers. This misunderstanding was particularly prevalent
during the early conceptualizations of sustainability. This is not an uncommon
phenomenon when a new paradigm emerges. However, as perspectives have begun to
converge, we are now seeing an increasingly uniform understanding and application of
the term sustainability as the triple bottom line.
Replacing this focus on standalone environmental issues have been articles which
take a more holistic approach, either by explicitly incorporating social responsibility
(the intersection of environmental and social performance) or sustainability – the triple
bottom line. While none of the articles in the 1991-2000 time period investigated their
topic of interest using these lenses, 18.75 percent of the articles in the 2001-2010 time
period used CSR and 25 percent of the articles in the 2001-2010 time period used
sustainability as the frameworks for their research.

Industry
Unlike our subject analysis, we do not find the dramatic trends and differences in the
industries which were selected for investigation in the first versus the second decades
of the 20-year time period. There has, however, been a large focus on consumer product
 industries and the transportation industry (Section B of Table III). Consumer products SSCM
firms were often chosen because they have been closest to the consumer and have often
been earlier adopters of certain environmental and social initiatives (Carter and Carter,
1998). Transportation carriers have been chosen by researchers who have focused on
safety issues and because they have a large carbon footprint that is more obvious
to many of their stakeholders. The majority of the industries studied, however, fall
under the category of “multi-industry,” which could include consumer products and 55
transportation carriers, but more often included multiple industries in manufacturing
or manufacturing together with service. This focus by researchers is understandable,
given the need to generate adequate sample sizes and perhaps the desire to increase the
external validity and generalizability of their findings.
However, this decision by researchers to use multi-industry samples also presents
an opportunity for future research to take deeper dives into individual industries as
sampling frames. More importantly, researchers should carefully select individual
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

industries with the goals of identifying specific types of sustainability activities that
are germane to those industries, and industries in which the boundaries of specific
theories might be extended or shown not to apply. As one broad example, researchers
are just beginning to examine service supply chains (Sampson and Spring, 2011),
defined here as the procurement and distribution of services (whether by
manufacturing or service industries). The sustainability characteristics of service
supply chains likely differ from those of their manufacturing counterparts and will be
ripe grounds for future investigations.

Theoretical lenses
The most striking result that is displayed in Section C of Table III is the relative dearth
in the use of a theoretical lens(es) to examine problems of interest in the sustainability
arena. About 55 percent of articles over the full-time period failed to employ any sort of
theory. It is encouraging to note, however, that there is a very strong trend toward
integrating theory in SSCM research. While more than 87 percent of articles did not
employ a theoretical lens in the 1991-2000 time period, this omission dropped to just
over 33 percent in the 2001-2010 time frame.
Of the theories which have been used, stakeholder theory (Freeman, 1984) has been
the most prevalent, followed by the resource-based view, and the closely aligned
natural resource-based view (Hart, 1995). The other category includes a diverse set of
theories from multiple fields, such as examining the role of dynamic capabilities in
achieving competitive advantage in sustainable global supplier management (Reuter
et al., 2010), using brand equity theory to explain how competitive advantages are
derived from sustainability (Flint and Golicic, 2009) and employing self efficacy theory
to better predict safe employee behavior (Brown et al., 2000).
Another trend which is displayed in Section C of Table III is the increased use over
time of multiple theoretical lenses within the same study. When done well, such
blending of diverse, complementary, and even overlapping theories can help to better
develop hypotheses, add rich insights to the interpretation of findings, and help better
understand the boundaries of where these theories apply. As one example, Pagell et al.
(2010) show that the traditional purchasing portfolio matrix developed by Kraljic is not
effective within the realm of SSCM. The authors integrate transaction cost economics,
the resource-based view, and stakeholder theory to modify the traditional purchasing
 IJPDLM portfolio matrix and create a theoretically derived purchasing portfolio matrix which
41,1 managers can use as a highly effective strategic tool.
Interestingly, transaction cost economics is one of the lesser used theories in the
SSCM literature that we reviewed. This suggests an opportunity for future research.
One particularly relevant facet of transaction cost economics is that of the bounded
rationality of actors, which occurs due to limits associated with communication and
56 information processing capabilities, and relatedly, the potential for opportunistic
behavior (Williamson, 1975, 1985). Bounded rationality also exists due to a lack of
transparency, and this can lead to both opportunistic behavior and perceptions of
opportunism (whether real or imagined). Such perceptions of opportunistic behavior
can be exacerbated as firms source from distant suppliers and as suppliers and other
members of the supply chain make claims regarding how their products are
manufactured (e.g. “organically certified,” “safe working conditions,” “low emissions,”
etc.). Future research might examine how bounded rationality and perceptions of
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

opportunism within the context of SSCM impact the decision to source domestically or
even locally, as opposed to internationally, and how supply chain governance
structures are affected – e.g. a muscular versus benign approach to hybrid contracting
(Williamson, 2008).

Methodology and analysis

Validity. Validity is considered to be the “sine qua none” of empirical research. Indeed,
without a rigorous examination of validity, the results of empirical studies are at best
suspect. Disturbingly, over 42 percent of the papers did not address validity and
another 11.7 percent only partially addressed empirical validity for the 1991-2010 time
period (Section D of Table III). However, similar to our other findings, we can see a
positive trend concerning methodical assessments of validity in the articles included in
our review. Over 64 percent of articles rigorously assessed validity in the 2001-2010
time period, versus fewer than 19 percent in the 1991-2000 time period. In the future,
authors must assess not only reliability but also the multiple facets of validity which
are relevant to their empirical methodology(ies), including content validity, convergent
validity, and discriminant validity.
Social desirability bias. Where applicable, articles were coded as either addressing or
not addressing social desirability bias[2]. This bias refers to the tendency of study
participants to provide answers and share perspectives that they believe will be viewed
favorably by the researcher (Crowne and Marlowe, 1960). Over the full-time period,
only 15.4 percent of the applicable articles addressed social desirability bias (Section E
of Table III). It is encouraging to note that this percentage increased from a scant
3.45 percent in the 1991-2000 time period to 25 percent in the 2001-2010 time period.
Still, a large majority of the papers did not assess social desirability bias, or even
discuss the potential limitations associated with this bias. While there may be
pragmatic limitations to assessing this sort of bias in some studies (ethnographies and
even certain case study designs come to mind), there are abbreviated versions of the
Crowne-Marlow social desirability bias scale which might be employed (Carter and
Jennings, 2004) for survey and laboratory study methodologies in future research.
Unit of analysis. The most common unit of analysis displayed in Section F of
Table III is the firm, followed by the individual, and the function or group (e.g. the
buying center). Unlike many of the other aspects of article methodology and analysis,
 we did not find large trends or shifts in the unit of analysis across time periods. There SSCM
has, however, been an increased emphasis on the supply chain as the unit of analysis
(from no articles in the first time period to 6.67 percent in the second time period) and
other units of analysis such as a project or a published article (again an increase from 0
to 6.67 percent across the two time frames).
Given the relatively large emphasis on the firm as the unit of analysis, our
findings suggest at least three viable opportunities for future SSCM research. First, 57
there is an opportunity to use the individual as the unit of analysis. While we are
beginning to develop an understanding of what drives firm behavior, we have much less
of an understanding of the drivers of individual managers’ behavior and of their
decision-making processes (Gattiker and Carter, 2010; Kaufmann et al., 2011).
The concept of behavioral supply management, defined as, “the study of how judgment
in supply management decision-making deviates from the assumptions of homo
economicus” (Carter et al., 2007, p. 634), would be a very logical and fruitful intersection
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

with SSCM. Some sample research questions could include examining supply chain
management employees as internal stakeholders, and how employee attitudes and
commitment to organizations might differ based on differing levels of SSCM (see
Mowday (1998) for an excellent synthesis of the organizational commitment literature);
examination of the biases that can enter the individual decision-making process, and
how these biases can impact the efficacy of SSCM initiatives; and investigation of how
individual managers can influence and gain the commitment of key internal
stakeholders to bring SSCM projects to fruition.
Second, the supply chain, even at the dyadic level, is consistently underrepresented as
a unit of analysis. While the collection of even dyadic data (as opposed to data involving
three or more firms) is quite difficult using methodologies that we have traditionally
relied upon (e.g. surveys and case studies), there are certainly opportunities to use
archival and secondary data to investigate SSCM. Calantone and Vickery (2010) provide
a number of suggestions concerning archival data sources that might be applied to
supply chain problems in general. Some of these data sources, such as Compustat, might
be combined with other secondary data sources, such as the Wharton Research Data
Service, the Dow Jones Sustainability Indexes, and the Global Reporting Initiativee to
investigate, for example, the relationship between company environmental and social
performance versus economic performance, and the relationship between regulatory
compliance and economic performance across members of a supply chain. Further, most
of these databases allow researchers to employ “time series, longitudinal panels, or time
series cross-sectional designs,” (Calantone and Vickery, 2010, p. 5) to complement and
extend existing findings (Carter et al., 2000; Klassen and McLaughlin, 1996).
Methodology. About 60 percent of the articles included in our systematic review of
the literature used a survey as the primary methodology to collect data (Section G of
Table III). There are, however, encouraging trends relating to the dominance of this
methodology: while almost 80 percent of the articles from the 1991 to 2000 time period
used a survey methodology, this percentage dropped to , 50 percent in the later,
2001-2010 time frame. Table III shows a sharp increase in the use of case studies,
archival data, and individual interviews during the second half of our 20-year period of
study. This is an encouraging sign, since the use of multiple methodologies allows for
triangulation and the maximization of internal validity, external validity, and realism
across studies (McGrath, 1982). In addition, we will probably see a further decline in
 IJPDLM the use of survey methodologies in the future, due to the difficulty of collecting large
41,1 sample sizes from cross-sectional surveys and limitations associated with common
method variance. We will, however, likely see surveys used at least somewhat more
often within individual organizations or dyads.
One troubling gap in the data displayed in Table III is the lack of laboratory
and field (quasi) experiments. In addition, there are opportunities to use richer
58 qualitative approaches such as ethnographies (see the work of Belk et al. (1988, 1989)
for an example of an innovative approach to using ethnographies within the field of
marketing). Both of these approaches can help to better triangulate findings across
extant studies.
Analysis. The body of empirical SSCM research relied heavily on descriptive
statistics during the first half of the past 20 years (54.29 percent, as given in Section H
of Table III); this percentage dropped to below 12 during the 2001-2010 time period.
This progression appears to follow a similar movement across the broader body of
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

empirical research in supply chain management (Carter and Ellram, 2003). This is a
very encouraging trend for two reasons. First, these summary statistics often lack
richness in terms of both advancing theory and providing value-adding managerial
implications. Second, and somewhat ironically, while “unsophisticated” from a
statistical standpoint, many of the summary statistics which employed means testing
were incorrectly performed. In particular, many of these analyses involved a large
number of univariate means tests, which failed to adjust individual p-values and thus
suffered from alpha inflation. In other words, many of the early analyses which
involved means testing likely reported significant differences when in fact, some of
these significant differences did not exist.
Often replacing the use of summary statistics, papers published in the 2001-2010
time period utilized regression analysis, structural equation modeling, path analysis,
and rigorous qualitative data analysis to investigate inferential relationships, and
factor analysis to assess measurement validity. In addition, while only about 10 percent
of the articles during the full 20-year time period included moderating variables in their
models and/or tested for interaction effects (Section I of Table III), this percentage
increased from 6.25 percent during the first half of the time period to over 13 percent
during the 2001-2010 time frame.
We have also begun to see more innovative approaches to traditional analyses.
For example, one particularly innovative alternative to analyzing qualitative data using
content analysis was recently employed by Tate et al. (2010). Here, the authors used
center resonance analysis to examine the CSR reports of 100 leading sustainability
organizations. Their analysis allowed them to develop rich understandings of how
these companies integrate the triple bottom line into both their internal operations and
their broader supply chains. We have also begun to see the use of social network
analysis to analyze the relationships among actors as the unit of analysis, as opposed to
the actor as the more traditional unit of analysis (Carter et al., 2007). In addition, it is
likely that we will see the use of more advanced econometric analyses as researchers
begin to employ archival data to test theoretically derived models (Calantone and
Vickery, 2010).
Finally, , 4 percent of the articles in our study used conceptual theory building as a
methodology to develop or expand theoretical insights. The field of supply chain
management has generally relied upon theories developed in adjacent
 fields (e.g. management – the resource-based view, institutional theory – and SSCM
economics – transaction cost economics). This is also true within the realm of SSCM.
Given the nascency of sustainability as a whole, along with the pivotal role that supply
chain management can play in an organization’s sustainability efforts, the area of
SSCM is ripe for such theory development. While Carter and Rogers (2008) provide an
initial theory development effort, much additional work is needed.
One likely reason for the lack of conceptual theory development in SSCM is a lack of 59
methodological training in the conceptual theory development methodology among the
broader community of supply chain management scholars. This lack of training likely
limits both motivations to use conceptual theory building and the effectiveness of
actual efforts. Further, when journal reviewers lack this training, this can act as an
additional barrier to publishing such research. We refer the reader to a forthcoming
discussion forum in the Journal of Supply Chain Management (2011, Vol. 47 No. 2),
which will address these issues and offer guidance concerning both how to craft and
Downloaded by RMIT University At 03:46 07 January 2016 (PT)

review conceptual theory development papers.

Conclusion
We must admit that many of the trends that appear in the individual analyses of our
data are probably signs of a broader trend toward performing more rigorous research
in the field of supply chain management as a whole. Still, we believe that these
individual analyses, and the trends which they reveal, help to further highlight the
many exciting research opportunities in the area of SSCM. As noted in the paper’s
introduction, there are several reasons why SSCM is enduring and not simply the
“flavor of the month.” The broad concept of sustainability, and the key interfaces that
sustainability has with supply chain management, strongly suggests that
sustainability is instead license to do business in the twenty-first century.
And supply chain management is an integral component of this license. Our hope is
that our systematic review of the literature, coupled with our own perspectives and
experiences, will help to meaningfully guide future research in this strategic and
imperative area of supply chain management.
Notes
1. n ¼ 80 is a relatively large number of units of analysis compared to most qualitative data,
which are analyzed in our field.
2. For certain methodologies and data sources, such as conceptual theory building and archival
data analysis, this field was coded as “not applicable”.

References
Belk, R.W., Sherry, J.F. Jr and Wallendorf, M. (1988), “A naturalistic inquiry into buyer and seller
behavior at a swap meet”, Journal of Consumer Research, Vol. 14 No. 4, pp. 449-70.
Belk, R.W., Wallendorf, M. and Sherry, J.F. Jr (1989), “The sacred and the profane in consumer
behavior: theodicy on the odyssey”, Journal of Consumer Research, Vol. 16 No. 1, pp. 1-38.
Brown, K.A., Willis, P.G. and Prussia, G.E. (2000), “Predicting safe employee behavior in the steel
industry: development and test of a sociotechnial model”, Journal of Operations
Management, Vol. 18 No. 4, pp. 445-65.
Carroll, A.B. (1979), “A three-dimensional conceptual model of corporate social performance”,
Academy of Management Review, Vol. 4 No. 4, pp. 497-505.