Journal of Operations Management 32 (2014) 446461

Contents lists available at ScienceDirect

Journal of Operations Management

journal homepage: www.elsevier.com/locate/jom

Developing supplier integration capabilities for sustainable

competitive advantage: A dynamic capabilities approach
Evelyne Vanpoucke a, , Ann Vereecke b , Martin Wetzels a
a
Marketing and Supply Chain Management Department, School of Business and Economics, Maastricht University, PO Box 616, 6200 MD Maastricht,
The Netherlands
b
Operations and Technology Management Center, Vlerick Business School and Ghent University, Reep 1, 9000 Gent, Belgium

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

Article history: Previous research describes supplier integration as a competitive resource that manufacturers use to
Available online 16 September 2014 create economic rents. Considering the mixed results obtained from linking supplier integration with
performance outcomes, a dynamic component or the ability to recongure the supply chain to adapt
Keywords: to changing environments appears critical to creating a sustainable competitive advantage. This study
Dynamic capability identies integration sensing, seizing and transforming as sub-capabilities that together form a dynamic
Buyersupplier integration
capability, referred to herein as supplier integrative capability (SIC). That is, SIC enables buyers to sense
Operational performance
changes in the supply environment by sharing information with suppliers, seize opportunities presented
Dynamics
Supply base complexity
by establishing procedures to analyse this information and make long-term changes to existing processes.
A global sample from the industrial sector reveals that the three capabilities exhibit complementarity and
must exist simultaneously for the capability to be effective, which then enhances both process exibility
and cost efciency and helps rms avoid the traditional trade-off of cost and exibility. In addition,
market and technological dynamics strengthen the effect of SIC on operational performance; supply base
complexity attenuates this link.
2014 Elsevier B.V. All rights reserved.

1. Introduction
As supply chain managers scrape their cost barrels, looking for
costs to eliminate, companies seek more innovative models that
can not only reduce costs but also improve services. Such innova-
tive models often are inspired by changes, customer requests and
opportunities in the market that require rms to renew themselves
by altering their resources and competences over time. Accord-
ingly, supply chain managers must adapt their integration practices
to changes in the market by exploiting their dynamic capabilities
(Teece, 2007; Teece et al., 1997) to renew their supplier integra-
tion practices and change their resources (processes or products)
(Danneels, 2011). As indicated by Jarrat (2004), all well-designed
relationship practices require continuous adaptations.
In practice, 61% of manufacturing companies worldwide
acknowledge that supply chain integration can contribute to their
prots by facilitating new waves of transformational processes in
Corresponding author. Tel.: +0032 485 063 663.
E-mail addresses: e.vanpoucke@maastrichtuniversity.nl (E. Vanpoucke),
Ann.vereecke@vlerick.com (A. Vereecke), m.wetzels@maastrichtuniversity.nl
(M. Wetzels).
the supply chain (Underwood and Agg, 2012). For example, the
head of value chain management at Novartis explains that to stay
ahead of competition, the vaccine rm must share information
with suppliers; set up systems and procedures to create smooth
supply ows, which in turn help resolve every day supply prob-
lems; and ensure that buyers and suppliers both develop innovative
supply chain projects to support long-term, cooperative objectives
(Deshais, 2012). Moreover, as stated by the Director of Integrated
Logistics at Kuehne and Nagel: Businesses cannot simply sit back
on their laurels and organizations will need to constantly evalu-
ate and adjust their supply chain operations. The challenge is that
it can become a rather perpetual and cyclical process that seems
to require constant re-evaluation and change as the ground moves
beneath it (Underwood and Agg, 2012).
But not all rms are equally successful in setting up such capa-
bilities. Whereas Frohlich and Westbrook (2001) argue that higher
levels of inter-organizational integration practices lead to better
operational and rm performance, other studies suggest negative
effects of supplier integration (Anderson and Jap, 2005; Villena
et al., 2011). Because some rms fail to benet from integration
practices, while others successfully manage supplier integration to
create value (Kale et al., 2002), we seek to determine which factors
lead to successful supplier integration and posit that differences in

http://dx.doi.org/10.1016/j.jom.2014.09.004
0272-6963/ 2014 Elsevier B.V. All rights reserved.
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
performance improvements might depend on the extent to which
a buyer can adapt its supply chain to new realities.
In particular, research has shown that using suppliers exper-
tise and knowledge affects a buyers ability to learn, adapt and
create change (e.g., Clark and Fujimoto, 1991; Jarrat, 2004; Kim
et al., 2011; Ragatz et al., 2002). This learning process can be stim-
ulated by creating integrative capabilities, such as processes to (1)
exchange information, (2) analyse the information and (3) adapt
the supply chain to cope with the new realities. Accordingly, we
propose that companies can set up successful supplier integra-
tive capabilities (SIC) that enable them to improve operational and,
ultimately, nancial performance. Moreover, three reasons might
explain why companies fail to develop SIC: (1) buyers and sup-
pliers fail to exchange the necessary information; (2) even if they
share this information, they may have difculties systematically
analysing it to understand what is happening in the supply chain
and (3) even if they share and analyse the information, they may not
possess the capability to exploit the information and recongure
their supply chain to respond to environmental challenges.
Our research thus contributes to supplier integration literature
in four ways. First, we conceptualize and test the notion that sup-
plier integration constitutes a dynamic capability, related to success
in buyersupplier relationships. Second, using survey data from a
global sample in the industrial sector, we show that rms with
SIC, including all three sub-capabilities of integration sensing, seiz-
ing and transforming, can increase their operational and nancial
performance. Third, we describe how market and technological
dynamics, as well as supply chain complexity, moderate the rela-
tionship between SIC and performance. Fourth, we apply dynamic
capabilities to understand why some rms enjoy greater benets
from supply chain integration and thus build a foundation for inter-
disciplinary research.
Starting from the resource-based view, we introduce the idea
of a dynamic capability and apply it to the notion of supplier inte-
grative capability. Next we formulate our hypotheses and discuss
our research model in Section 3, then present the results of the
empirical hypotheses tests in Section 4. Finally, we offer discuss-
ions of the implications of the results for OM researchers (Section
5) and practitioners (Section 6), as well as some limitations and
recommendations for further research (Section 7).
2. Theoretical framework and hypotheses development
2.1. From the RBV to dynamic capabilities
The resource-based view of the rm (RBV) is a theoreti-
cal framework that seeks to explicate how rms, as bundles
of heterogeneous resources that are valuable, rare, inimitable
and non-substitutable, achieve sustainable competitive advantages
(Barney, 1991; Penrose, 1959; Peteraf, 1993). The value of the
RBV lies in its ability to identify which resources dene a rms
success. A capability view complements the RBV by identifying
which capabilities help rms apply their resources across mul-
tiple environments or situations. According to a Schumpeterian
(1950) perspective, it is difcult to maintain sustainable compet-
itive advantages in dynamic environments, so rms constantly
must recongure their resources to t changing situations. In turn
rms need dynamic capabilities that enable them to create, extend
and modify the ways they earn their living (Helfat et al., 2007).
Similar to Teece et al. (1997) and Eisenhardt and Martin (2000,
p. 1107), we dene dynamic capabilities as the rms process
to use resources specically, to integrate, recongure, gain and
release resources to match and even induce market change. With
these organizational and strategic practices, rms achieve new
resource congurations as their markets evolve. In turn, dynamic
447
capabilities drive the creation, evolution and recombination of
other resources into new sources of competitive advantage. In
contrast to operational capabilities, which enable a rm to earn
a living in the short term and maintain the status quo (Helfat
et al., 2007), dynamic capabilities help it constantly renew oper-
ational capabilities (Winter, 2003) and address long-term changes
in its environment. Helfat and Winter (2011) also caution that the
line between operational and dynamic capabilities is unavoidably
blurry, because change is always occurring to some extent. Dis-
tinctions such as new versus existing business cannot differentiate
between operational and dynamic capabilities, and some capabili-
ties even serve dual purposes (Helfat and Winter, 2011). Even with
these overlaps though, operational and dynamic capabilities sup-
port distinct purposes, i.e., short-term and long-term performance
improvements (Helfat et al., 2007; Winter, 2003).
2.2. Dening SIC as a dynamic capability
Some dynamic capabilities enable rms to enter new business or
create new products or processes (Helfat et al., 2007). Across these
types of dynamic capabilities, information processing capabilities
consistently serve to help the rm identify the nature of the chang-
ing environment and thus sense opportunities (Helfat et al., 2007;
Pierce et al., 2002). We focus on supplier integrative capabilities
(Helfat et al., 2007) as essential routes to learning from suppliers
and adapting the supply chain to changes in supply.
Integration refers to the unied control of a number of suc-
cessive or similar economic or especially industrial processes
formerly carried out independently (Webster, 1966, p. 1175). In
a buyersupplier context, supplier integration is the degree to
which a manufacturer partners with its suppliers to structure
inter-organizational strategies, practices and processes into collab-
orative, synchronized processes (Droge et al., 2012; Flynn et al.,
2010; Katunzi, 2011). Swink et al. (2007) denes operational inte-
gration as the coordination of daily ows, such as transactions,
material movements and ordering processes, to achieve effective
movements of products, services, information, money and deci-
sions, which in turn provides maximum value to the customer at
low cost and high speed (Flynn et al., 2010; Frohlich and Westbrook,
2001).
Starting from these denitions, we dene supplier integra-
tive capability as a dynamic capability that contains processes
to achieve effective and efcient product and information ows
between buyers and suppliers (Carr and Pearson, 1999; Swink et al.,
2007), as well as the ability to adapt these processes to environmen-
tal change. Then, in line with previous research, we argue that SIC
serves dual purposes: It enables communication and coordination
across organizational units and rms (Fortune and Mitchell, 2012),
such that it can facilitate smooth operations and delivery processes,
and it supports dynamic purposes, such as introducing a new dis-
tribution channel or sourcing from different regions to mitigate
supply risks. Thus SIC might be dynamic and operational, depend-
ing on its nature and intended use (Helfat and Winter, 2011). Most
extant literature addresses operational goals (e.g., Frohlich and
Westbrook, 2001; Vereecke and Muylle, 2006); we seek to add a
dynamic view.
2.3. Processes underlying SIC
Teece (2007) identies three core processes of a dynamic
capability: (1) sensing, (2) seizing and (3) transforming. These
sub-capabilities are complementary, such that only when they
are combined and aligned can they produce a dynamic capabil-
ity. Transforming, which is responsible for modifying integrative
processes, particularly ensures the dynamic nature of the capa-
bility, in that it is the ability to adapt processes to changing
448 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
environments. Using this framework, we assert that SIC combines
the three distinct sub-capabilities of integration sensing, seizing
and transforming.
The sensing capability entails processes for gathering and inter-
preting data, allocating resources and tasks and communicating
decisions and information (Sanchez and Heene, 1996). It is acti-
vated by the presence of systems to scan, learn, lter, interpret
and calibrate data and information for identifying opportunities
and threats. These searches occur at the core and the periphery
of the companys business system and should embrace exist-
ing and potential collaborators, such as suppliers (Teece, 2007).
In buyersupplier relationships, an integration sensing capability
results from information sharing practices that inform partners
about current and future physical ows (Barratt and Oke, 2007;
Kulp et al., 2004; Mller and Gaudig, 2011). Zhou and Benton
(2007) show that such integration sensing signicantly enhances
supply chain practices (i.e., seizing). The information exchanged
might include inventory, replenishment, planning, forecasting and
tracking data (Kulp et al., 2004; Monczka et al., 1998). Such crit-
ical, often proprietary, information might get exchanged through
face-to-face meetings, telephone, fax, mail and the Internet (Mohr
and Spekman, 1994). A successful information exchange should
be frequent, bidirectional, informal and noncoercive (Mohr and
Spekman, 1994; Mller and Gaudig, 2011; Vanpoucke et al., 2009),
so that it reduces bullwhip effects (e.g., Cachon and Fisher, 2000;
Lee et al., 2000) and supply chain costs (e.g., Droge et al., 2012; Flynn
et al., 2010). Then, by using the shared information, each partner
can make better decisions about ordering, capacity allocation and
production/material planning to optimize supply chain dynamics.
Access to this information enables the buyer to sense what is hap-
pening in the supply chain and determine when and how to react by
making changes to the supply chain (Gunasekaran and Ngai, 2004;
Harland et al., 2007).
Seizing includes enterprise structures and procedures for identi-
fying threats and opportunities (Teece, 2007). Specically, it helps
rms construct decision-making protocols, build loyalty and com-
mitment and control the environment. In a buyersupplier context,
collaborative planning helps buyers seize the supply chain through
collaborations among partners to develop and execute inven-
tory and replenishment plans, as exemplied by vendor-managed
inventory (VMI) (Govindan, 2013) and just-in-time replenishment
systems (Lage and Filho, 2010). These systems reveal future contin-
gencies and the resulting relationship duties and responsibilities
(Claro et al., 2003). Such collaborative planning can complement
integration sensing activities (Foerstl et al., 2013); it also offers a
more comprehensive picture for managing the supply chain. Inte-
gration seizing practices also might attempt to streamline supply
chain activities, such as when a supplier customizes its produc-
tion and transportation systems by synchronising these processes
with those of its buyer. Such collaborative procedures and actions
relate positively to supply outcomes (e.g., Flynn et al., 2010; Kulp
et al., 2004; Monczka et al., 1998). As Cai et al. (2010), Devaraj et al.
(2007) and Kulp et al. (2004) note, information sharing is neces-
sary to understand the supply chain, though it cannot enhance
the competitiveness of the supply chain if the members lack the
means to coordinate their planning. Instead, rms need to apply
external knowledge to their planning decisions, anticipate deliv-
ery problems and proactively manage opportunities and threats to
their physical ows. Integration seizing is a necessary element of
integrative capabilities, because rms need structured processes
in place to process data to be able to understand it and then share
their interpretations of the data with their partners (e.g., Smaros,
2007; Zhou and Benton, 2007).
The third sub-capability, transformation, is the continuous align-
ment and realignment of operational practices (Teece, 2007).
Transforming capabilities change current processes in a dynamic
environment. In buyersupplier relationships, transformation
capabilities might be embedded in supplier development projects
(Krause et al., 2007; Wagner and Krause, 2009) or additional invest-
ments in coordination systems, which enable rms to learn from
one another, co-specialize, make changes to their own internal
processes when needed and continuously improve supply chain
processes. Such transformation processes are in line with Swink
et al.s (2007) concept of strategic integration, entailing activi-
ties that acquire, share and consolidate strategic knowledge with
partners outside the organization to align and t with business
processes. These activities are especially important in moderately
(Eisenhardt and Martin, 2000) and highly (Amit and Schoemaker,
1993) dynamic environments. The combination of integration
sensing and seizing capabilities allows rms to determine when
opportunities arise and exploit them in the short and medium
term. Transformation is needed to cope with long-term changes
that require strategic action. That is, it is the addition of the trans-
formation capability that makes an integrative capability dynamic.
Research has considered how individual sub-capabilities or
some pairing relates to performance (e.g., Vereecke and Muylle,
2006; Zhou and Benton, 2007), but as we argue, the complemen-
tarities among all three sub-capabilities cause them to form the
dynamic capability of SIC. With SIC, companies can create a sus-
tainable competitive advantage by seeing what is happening in the
supply chain, understanding the causes and the circumstances of
the current situation and then transforming their supply chains by
combining current resources and knowledge to cope with the cur-
rent sensed and seized challenges. This article offers an empirical
test of this theory of complementarity in a supply chain context, as
described in Fig. 1.
Teeces (2007) argument for the simultaneous presence of the
three capabilities also is consistent with Venkatramans (1989)
notion of t as co-alignment and Edwardss (2001) integrative ana-
lytical framework for multidimensional constructs, in which an
integrative capability is a higher-level construct (Edwards, 2001).
Therefore, we test whether SIC is a dynamic capability by modelling
it as a co-alignment construct that consists of integration sensing,
seizing and transforming sub-capabilities (Fig. 1).
2.4. Impact of SIC on rm performance
The impact of dynamic capabilities on rm performance is a
central but as yet unresolved issue among strategic scholars. To
explore the link between SIC and performance, we turn to opera-
tions management literature (Rosenzweig et al., 2003) and propose
that competitive advantages do not stem directly from SIC but
rather move through its impact on operational performance. As
in Fig. 1, we assume an indirect impact of SIC, through operational
performance, on rm performance.
Operational performance refers to a manufacturers actual or
realized competitive strength (Rosenzweig et al., 2003). Some
research indicates positive effects of integration on operational
performance (e.g., Droge et al., 2012; Frohlich and Westbrook,
2001; Jacobs et al., 2007; Kulp et al., 2004), though other studies
either nd no signicant link or indicate negative inuences (e.g.,
Rodrigues et al., 2004; Swink et al., 2007; Vereecke and Muylle,
2006). These inconsistencies might reect different supplier inte-
gration practices or operational performance measures, as well as
complementarities across supplier integration practices (Alfalla-
Luque et al., 2013).
We argue that cost and exibility are two opposing dimen-
sions of operational performance (Boyer and Lewis, 2002; Sazadeh
et al., 2000), associated with distinct structural and infrastruc-
tural choices (Kotha and Orne, 1989; Sazadeh et al., 2000;
Skinner, 1974) and corresponding to different market strate-
gies (Miller and Roth, 1994). Fisher (1997) thus differentiates
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
Fig. 1. Supplier integrative capability model.
functional or cost-efcient supply chains from innovative or exi-
ble ones. Narasimhan et al. (2008) identify two main supply chain
management strategies: those focused on variability in the supply
chain and those that address the velocity and speed of informa-
tion and decision making. Supplier integration practices directed
toward increasing velocity correlate with nancial performance
through cost reduction practices, whereas variability and the ability
to respond to changing demands enhance process exibility prac-
tices. We assert that this process exibility enables rms to serve
customers better, which should lead to higher customer satisfac-
tion and thereby increase sales and market share (i.e., the rms
market performance). Kotha and Orne (1989) and Braunscheidel
and Suresh (2009) nd that integration can help develop exi-
ble operations; Rosenzweig et al. (2003) similarly contend that
tight integration across supply chain entities is a requirement for
developing process exibility and can lead to improved market
performance. We predict:
H1a. Supplier integrative capability positively affects process
exibility.
H1b. Supplier integrative capability positively affects cost ef-
ciency.
H2. Operational performance (process exibility and cost ef-
ciency) mediates the relationship between supplier integrative
capability and rm (market and nancial) performance
2.5. Impacts of environmental dynamics and supply complexity
Teece (2007) argues that the possession of dynamic capabilities
is especially relevant to rms in business environments marked by
dynamism and complexity. In such settings, success depends not
on the rms ability to engage in optimization or capture existing
economies of scale but rather its capacity to transfer know-how
efciently and effectively, upgrade its best practice business pro-
cesses and invent new business strategies. Firm success depends
on the ability to cope with change in complex business environ-
ments. Accordingly, we suggest two categories of moderators in the
449
relationship between SIC and operational performance: external
dynamics and supply complexity.
2.5.1. External dynamics
In environments characterized by rapid market and techno-
logical change, rms confront uctuations, unpredictable events
and discontinuities that may erode their operational performance,
unless they have the dynamic capability to adapt to changes
(Protogerou et al., 2011). Disruptive events, such as the introduc-
tion of a new technology, new supply channel or drastic market
change, require dynamic capabilities to recongure operational
processes and support the continuous development of new supply
chain solutions. Dynamic capabilities vary with market and techno-
logical dynamics (Eisenhardt and Martin, 2000; Peteraf et al., 2013).
When markets are moderately dynamic, dynamic capabilities are
embedded in complicated, detailed practices, whereas in highly
dynamic environments, they are embedded in simple, experien-
tial, unstable processes. Eisenhardt and Martin (2000) and Helfat
and Winter (2011) also note that even capabilities that support
existing businesses or seemingly non-radical change may have
important dynamic attributes, because most change occurs in grad-
ual fashion. Even if the extent of change is limited, there is a
need to adapt or continuously improve existing operational capa-
bilities to ensure they maintain their value. Therefore, dynamic
capabilities are not limited to dynamic environments; they can
offer value in less dynamic market and technology contexts, where
they constitute detailed, analytical processes for effecting incre-
mental change. Yet the impact of dynamic capabilities, including
SIC, should be stronger in more dynamic environments. Thus, we
hypothesize:
H3a. The impact of supplier integration capability on operational
performance (process exibility and cost efciency) grows stronger
as markets become more dynamic.
H3b. The impact of supplier integration capability on operational
performance (process exibility and cost efciency) grows stronger
as technologies become more dynamic.
450 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
2.5.2. Supply base complexity
Choi et al. (2001) argue that complexity naturally stems from the
extensive interconnectedness of supply networks, in which sup-
pliers connect to various supply chains that ultimately generate
diverse products and unanticipated customers. The complexity of
the supply base thus is driven by both the number of supplier rela-
tionships and the geographical area in which they locate; it also
tends to be associated with high product variety, small batches,
high expectations of operational performance, high risks and exten-
sive innovation of products and processes (Gimenez et al., 2012).
Adding suppliers increases complexity by expanding the num-
ber of information ows, physical ows and relationships to
manage (Bozarth et al., 2009). Many rms seek to manage their sup-
ply base actively through optimization or rationalization projects
(Choi and Krause, 2006), which attempt to reduce transaction costs
by concentrating purchase volumes with fewer suppliers. This step
often leads to immediate cost reductions, then enables rms to
focus on longer-term strategic issues, such as supplier development
or closer synchronization. That is, a smaller supply base makes sup-
plier integration easier and increases the value of SIC for operational
performance.
Supply base complexity also might increase as the supply base
expands globally. Global linkages expose rms to a wider range
of complications, including import laws, uctuations in currency
valuations, cultural differences and longer and more uncertain sup-
ply lead times (Cho and Kang, 2001). Long or unreliable supply
lead times can force rms to adopt planning and material manage-
ment processes with longer planning horizons and greater detail
(Vollmann et al., 2005). Chen et al. (2000) show that the magnitude
of the bullwhip effect is driven by replenishment lead times and
even can be expressed as a function of lead time. That is, the longer
the lead time, the more difcult it is to manage demand uctuations
in the supply chain. A global supply base also may imply cultural dif-
ferences that hinder knowledge exchanges. Extant research notes
the need for mutual understanding at the lower rm echelons,
where planners and purchasing assistants interact daily (Bierly and
Coombs, 2004), but cultural differences among these colleagues can
create misunderstanding (Simonin, 1999). Cultural disparities also
produce diverse expectations that might be detrimental to per-
formance outcomes (Cheung et al., 2011). In increasingly global
environments, more parties also are involved, with less informa-
tion available at any point in the production process, which makes
it much harder to identify, quantify, prioritize and mitigate risk.
Recent disasters in, for example, Japan and Thailand highlight the
need for effective risk management throughout the supply chain, to
help minimize disruptions and resume normal business conditions
quickly. Finally, global growth has increased the volume of regula-
tions that demand compliance (Savitz, 2012). Thus, we expect an
attenuated impact of supplier integration on performance for more
international supply bases and hypothesize:
H4a. The impact of supplier integrative capability on operational
performance (process exibility and cost efciency) is attenuated
as the number of key suppliers increases.
H4b. The impact of supplier integrative capability on operational
performance (process exibility and cost efciency) is attenuated
as the internationalization of the supply base increases.
3. Method
3.1. Sample
To test whether SIC improves performance, we deemed
a cross-sectional sample appropriate, as used previously in
strategic management literature to study dynamic capabilities
(e.g., Drnevich and Kriauciunas, 2011; Menguc and Auh, 2006;
Protogerou et al., 2011). Similar to Drnevich and Kriauciunas
(2011), we also included data about the current processes and
status of the rm, the rms transformation processes and its oper-
ational and performance improvements. Thus, we could capture
the dynamic nature of supplier integration and its impact on per-
formance.
Through membership on the design team of the International
Manufacturing Strategy Survey (IMSS), we accessed data collected
in the fth edition of the International Manufacturing Strategy Sur-
vey (IMSS-V; see www.manufacturingstrategy.net). This project,
originally launched in 1992 by the London Business School and
Chalmers University of Technology, studies the manufacturing and
supply chain strategies of manufacturers in the machinery and elec-
tronics industrial sectors (i.e., ISIC 2835 classications). A global
network of researchers uses a common questionnaire adminis-
tered simultaneously in many countries by local research groups
(Lindberg et al., 1998). The questionnaires were sent by ordinary
mail, fax or e-mail. To ensure reliable data gathering, the electronic
version of the questionnaire was preferred since the summary
Excel le automatically extracted data from each local, automated
questionnaire (Word document). In addition, all researchers were
requested to use fault-proof methods (e.g., double input) and the
Excel le contained controls for data consistency.
The data collection for IMSS-V started in January 2009 and ended
mid-2010. Considering our goals to obtain a large sample, analyse
moderately dynamic industries and keep manufacturing practices
relatively homogenous (i.e., the industrial sector, which is very dif-
ferent from other sectors like the process and the construction
sector), IMSS offers an appropriate data set.
The rst section of the questionnaire pertains to the business
unit (general information, such as company size, industry, pro-
duction network conguration, competitive strategy and business
performance); the other sections refer to the plant (manufacturing
strategies, practices and performance). Although the questionnaire
structure has remained the same for every edition, some ques-
tions have been updated or removed, and new questions have
been added by the international design team (Van de Vijver and
Leung, 1997). The original questionnaire is in English but also
has been translated by national researchers (e.g., for French- and
Spanish-speaking countries) using double- and reverse-translation
procedures, in a coordinated manner for countries with similar
languages.
The sample companies usually are selected from public or pri-
vate local databases based on ISIC code, and the point of contact is
the operations, production, supply chain or plant manager/director.
The selection process determines the knowledge and awareness
this manager exhibits toward both operational and strategic deci-
sions. If a respondent agreed to participate, the local researchers
sent a questionnaire; if necessary, they also provided reminders
after several weeks. This technique has been applied successfully
in previous research and helps increase response rates (Zhao et al.,
2008). The returned questionnaires were subjected to missing data
controls, handled on a case-by-case basis but usually by contacting
the company again. Every local research group also controlled the
gathered data for late respondent bias, company size and industry.
Finally, quality checks (e.g., checking for errors, missing data, out-
liers) were conducted, and all the data were grouped into a unique
database through central coordination by the Politecnico Di Milano
(Italy).
In total, 4272 questionnaires were distributed across the differ-
ent countries. After excluding cases with too much missing data or
many errors, the sample consisted of 719 rms from 20 countries
in America, Europe and Asia (see Table 1). The overall response rate
was 16.8%. Response rates for each individual country were above
10%, with the exceptions of the United Kingdom (5%), Korea (7%)
and Japan (9.3%).
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461 451

Table 1
IMSS-V sample prole.

Demographic dimension N Percentage (%)

Personnel employed
Small companies (<250 employees) 341 47.4
Medium companies (between 250 and 500 employees) 147 20.5
Large companies (>501 employees) 218 30.3
Missing 13 1.8
Total 719 100.0
Industrial sector
28 Fabricated metal products, except machinery and equipment 214 29.8
29 Machinery and equipment not elsewhere classied 206 28.7
30 Ofce, accounting and computing machinery 16 2.2
31 Electrical machinery and apparatus not elsewhere classied 83 11.6
32 Radio, television and communication equipment and apparatus 43 6
33 Medical, precision and optical instruments, watches and clocks 47 6.5
34 Motor vehicles, trailers and semi-trailers 52 7.2
35 Other transport equipment 37 5.1
Missing (but members of the ISIC 28-35 code) 21 2.9
Total 719 100.0
Geographical areas
Europe 414 57.6
Belgium 36
Denmark 18
Estonia 27
Germany 38
Hungary 71
Ireland 6
Italy 56
the Netherlands 51
Portugal 10
Spain 40
Switzerland 31
UK 30
North America 92 12.8
Canada 19
USA 73
South America 54 7.5
Brazil 37
Mexico 17
Asian countries 159 22.1
China 59
Japan 28
Korea 41
Taiwan 31
Total 719 100.0

To test for differences between respondents and non-
respondents, we considered demographic variables. Most local
researchers started from an existing database, with information
about all public rms in their country, so these readily accessible
secondary data could reveal any signicant differences between
respondents and non-respondents in their size, industry, sales or
proprietary structure. If such databases were not available, we
checked for non-response bias by using questionnaire items, such
as size, industry and manufacturing performance. In no cases did
we nd evidence of non-response bias.
3.2. Measures
In keeping with existing denitions of operational and dynamic
capabilities (Peng et al., 2008; Teece et al., 1997), we measure inte-
gration sensing, seizing and transforming capabilities as a bundle
of integration practices. This approach also is consistent with man-
agement theory, which suggests that practices and routines can be
integrated to reinforce one another (Milgrom and Roberts, 1995;
Peng et al., 2008). Practices embedded in systems offer more poten-
tial value than those being used alone (Peteraf, 1993). To leverage
the complementarities among practices and create specic capa-
bilities, strong performers often implement manufacturing and
administrative practices in bundles (Cua et al., 2001). Accordingly,
we measure capabilities as bundles of practices.
For integration sensing, we consider practices for learning and
sensing what is happening in the supply chain. Such learning
and sensing take place by sharing of different types of data and
knowledge with the supplier, including inventory levels, produc-
tion planning, demand forecasts and order tracking or tracing.
Similar scales are used as Wiengarten et al. (2014) and Vereecke
and Muylle (2006). Integration seizing practices instead refer to the
structures and procedures that help rms understand future prob-
lems, apply external knowledge to current planning decisions and
proactively manage opportunities and threats to physical ows.
Standardized supply chain procedures can support the manage-
ment of inventories at the buyers site (e.g., VMI, consignment
stock), the use of advanced replenishment methods and greater
physical integration by the supplier into the plant. Such practices
help the buyer coordinate and structure its relationship with sup-
pliers. We used similar scales as Wiengarten et al. (2014), Vereecke
and Muylle (2006) and Frohlich and Westbrook (2001). Finally,
integration transformation involves the continuous alignment and
realignment of specic assets to recongure supply chain assets
and facilitate long-term strategic changes. Many projects might t
this description, but supplier development and additional invest-
ments in coordination systems likely stimulate and continuously
improve integration in the supply chain. Therefore, we consider
the extent to which action programs in the past three years involve
supplier integration, such as rethinking and restructuring the
452 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461

Table 2
Descriptive statistics and bivariate correlations.

Variable Mean Std. dev. 1 2 3 4 5 6 7

Integration sensing 1 2.95 1.22 1

Integration seizing 2 2.15 1.10 0.67** 1
Integration transforming 3 2.81 1.07 0.48** 0.54** 1
Process exibility 4 3.01 1.12 0.14** 0.20** 0.30** 1
Cost efciency 5 2.58 1.01 0.13** 0.34** 0.31** 0.60** 1
Market performance 6 2.47 1.18 0.06 0.03 0.16** 0.38** 0.34** 1
Financial performance 7 2.15 1.22 0.02 0.03 0.11* 0.31** 0.31** 0.76** 1
*
p < 0.05.
**
p < 0.01.

supply chain, re-organizing the management of the supplier port- Table 3

Overall t indices.
folio, implementing supplier development opportunities or vendor
rating programs and increasing the level of coordination in plan- Fit indices: CFA Co-alignment Direct-effect
ning and ows, including dedicated investments. Each action model model
program enables coordination with suppliers to create longer-term, Chi-square 270.25 391.864 787.438
strategic change. Companies that have introduced change in sup- d.f. 131 143 141
plier base management in the past should possess a capability to RMSEA 0.038 0.045 0.080
CFI 0.972 0.951 0.872
transform the supplier base in the future.
IFI 0.973 0.951 0.873
In summary, we dene SIC as the rms ability to use sup- GFI 0.962 0.947 0.896
plier integration practices to respond to changes in the supply
environment by repurposing its supply chain resources. Com-
plementarities among the three processes of integration sensing, item. The descriptive statistics and bivariate correlations for all
seizing and transforming, as dened by the dynamic capabil- model variables appear in Table 2.
ity view, form the basis for stronger operating performance and
improved rm performance. 4. Results
To operationalize the constructs, we used multi-item, reective
rather than formative indicators (Bollen, 1989). Thus we could iden- With SEM, we can describe the unique inuences of each
tify items from the survey that correlated strongly with the latent integration process, controlling for spurious associations (Bollen,
variables in question, rather than requiring an exhaustive list of all 1989). As noted by Iacobucci et al. (2007), structural equation mod-
practices that might be related to these capabilities. This approach els are superior to regression when testing for mediation.
is important, because IMSS includes a nite number of practices The conrmatory factor analysis (CFA) and path models were
and routines related to integration capabilities. The items for each tested using EQS software version 6.1 (Bentler, 1992), according
construct used ve-point Likert scales. to the two-step approach recommended by Anderson and Gerbing
We also considered operational performance and rm per- (1988). That is, we assessed the rst-order measurement model to
formance as latent, multi-dimensional constructs. We measured investigate the validity of the constructs. As suggested by Fornell
two components of operational performance: improved cost ef- and Larcker (1981), we afrmed the validity and reliability of our
ciency and process exibility. Consistent with prior research measurement model before testing for relationships in the struc-
(e.g., Menor et al., 2007; Rosenzweig et al., 2003; Roth et al., tural model.
1992), we operationalized process exibility to comprise mix
exibility, volume exibility and customization elements. Cost 4.1. Measurement model
efciency included unit manufacturing, procurement and man-
ufacturing overhead costs. For rm performance, we considered The test of the measurement model relied on a CFA of all rst-
both market and nancial performance. Similar to Menor et al. order latent constructs; because we identify several integration
(2007) and Narasimhan and Kim (2002), our market perfor- practices relevant for describing a rms integration capabilities
mance measure relied on market share and sales growth. and used measurement scales from previous studies, CFA is an
Financial performance used return on sales and return on invest- appropriate technique (Shah and Goldstein, 2006). The CFA results
ment. indicated that the data t the measurement model well. We used
We included several control variables. For the size of the rm, incremental, absolute and parsimonious measures, each of which
we measured the log of the total number of employees, which cap- provides a different perspective on the match between the hypoth-
tured any effects of variations in rm size on performance (Im esized relationships, as estimated by the model, and the observed
and Rai, 2008). The log of the number of employees enabled us data (Hu and Bentler, 1995). We summarize the test statistics in
to evaluate the relative magnitude of resources available in the Table 3. The commonly used comparative t index (CFI) equalled
company for conducting integration initiatives. Consistent with 0.972 for our model. The root mean square error of approxima-
previous research, this variable should be useful in our global con- tion (RMSEA), an absolute measure of t, reached 0.038 (90%
text (McKone et al., 1999; Sila, 2007; Sousa and Voss, 2008). In condence interval [0.032, 0.045]). Finally, the parsimonious t
addition, we controlled for regional effects by checking for differ- measures (Normed chi-square) indicated the parsimony of the pro-
ences across continents. posed model, according to the model t relative to the number of
In the next section, we detail the reliability and validity of our estimated coefcients needed to achieve that level of t (Bentler,
measures and test the model using structural equation modelling 1992; Bollen, 1989).
(SEM). However, we rst tested whether the missing observa- Consistent with Anderson and Gerbing (1988), we used the
tions in our data set were random. According to Littles MCAR test CFA results to test for the uni-dimensionality of the latent con-
(2 = 1231.710, df = 1166, p < 0.05), they were. To handle the miss- structs. The path loadings were signicant and greater than 0.60,
ing values, we thus replaced them with the mean value for that with two exceptions (mix exibility = 0.57; order tracking with
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461 453

Table 4
Conrmative factor analysis (n = 631).

Measurement Standardized AVE Construct

item path loadinga reliability

Integration sensing 0.49 0.74

How do you coordinate planning decisions and ow of goods with your
key/strategic suppliers? (Scales 15: Level of Adoption: (1) none, (5) high)
x1 Share inventory level information with 0.79
suppliers
x2 Share production planning decisions 0.73
and demand forecast knowledge with
supplliers
x3 Order tracking/tracing with supplier 0.55
Integration seizing 0.45 0.71
How do you coordinate planning decisions and ow of goods with your
key/strategic Suppliers? (Scales 15: Level of Adoption: (1) none, (5) high)
x4 Vendor managed inventory or 0.68
consignment stock
x5 Just-in-time replenishment 0.70
x6 Physical integration of the supplier 0.62
into the plant
Integration transforming 0.53 0.77
Indicate degree of the following action programmes undertaken over the last
three years? (Scales 1-5: (1) none, (5) high)
x9 Rethinking and restructuring supply 0.67
chain strategy and the organization
and management of supplier portfolio
through e.g. tiered networks, bundled
outsourcing sourcing and supply base
reduction
x10 Implementing supplier development 0.73
and vendor rating programs
x11 Increasing the level of coordination of 0.77
planning decisions and ow of goods
with suppliers including dedicated
investments (in e.g. extranet/EDI
systems, dedicated
capabity/tools/equipment, dedicated
workforce, etc.)
How has your operational performance changed over the last three years?
(Scales 15: Compared to three years ago, the indicator has: (1) deteriorated
more than 10%, 2) stayed about the same, (3) improved 10 to 30%, (4)
improved 30 to 50%, (5) improved more than 50%)
Flexibility 0.55 0.82
y1 Mix exibility 0.57
y2 Volume exibility 0.78
y3 Product customization ability 0.85
Cost efciency 0.54 0.78
y4 Manufacturing costs 0.77
y5 Procurement costs 0.76
y6 Manufacturing overhead costs 0.66
What is your current business unit performance? (Scales 15: Compared to
three years ago, the indicator has: (1) deteriorated more than 10%, (2) stayed
about the same, (3) improved 10 to 30%, (4) improved 30 to 50%, (5) improved
more than 50%)
Market performance 0.64 0.78
y7 Sales 0.80
y8 Market Share 0.80
Financial performance 0.84 0.91
y9 Return on Sales (ROS) 0.92
y10 Return on Investment (ROI) 0.91

All of the path loadings are signicant at the 0.01 level.

supplier = 0.55; Table 4). We chose to retain these two items,
because they have appeared in previous research and are important
for our constructs. Furthermore, the measurement model showed
good overall t, so we can conclude that the constructs exhibited
uni-dimensionality. In addition, in support of convergent valid-
ity (Anderson and Gerbing, 1988), the manifest variables loaded
signicantly on their respective latent variables.
We also used the CFA to establish discriminant validity. For all
possible pairs of latent constructs, we rst allowed it to correlate
freely, then constrained the correlation between the two con-
structs to 1.00. The 2 differences between the unconstrained and
constrained models were all statistically signicant (p < 0.01), so the
constructs represented distinct scales (OLeary-Kelly and Vokurka,
1998). We also determined that the average variance extracted
(AVE) of each pair of factors was greater than their squared cor-
relation.
Finally, in terms of the reliability of the items and constructs
comprising the measurement model, we found that the com-
posite reliability and AVE statistics were acceptable (Fornell and
Larcker, 1981). The composite reliabilities ranged between 0.71
and 0.91 (Table 4), well above the 0.60 threshold, so they indicated
acceptable reliability levels for the constructs (Bollen, 1989; Fornell
454 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
Fig. 2. Supplier integrative capability structural model.
and Larcker, 1981). The AVE statistics spanned from 0.45 to 0.84
(Table 4), and we considered a level of 0.40 acceptable (Hatcher,
2003).
4.2. Structural model
To test the hypotheses, we rst analysed the t of the sam-
ple data to the proposed co-alignment model using SEM. The
overall t indices indicated that the co-alignment model t the
data well (2 = 391.86, df = 143, p < 0.01, CFI = 0.951, goodness-of-t
index [GFI] = 0.947, RMSEA = 0.045; 90% condence interval [0.039,
0.055]). According to the structural model and standardized path
coefcients in Fig. 2, SIC related positively to the three rst-order
constructs: sensing ( = 0.35, p < 0.01), seizing ( = 0.46, p < 0.01)
and transforming ( = 0.49, p < 0.01). Moreover, the prediction that
the three sub-capabilities are equally important for SIC was not
rejected, in initial support of the presence of SIC.
In support of H1a and H1b, SIC related strongly and directly to
both the process exibility ( = 0.69, p < 0.01) and cost efciency
( = 0.77, p < 0.01) components of operational performance. Each
component related signicantly to rm performance outcomes,
such that process exibility ( = 0.27, p < 0.01) and cost efciency
( = 0.19, p < 0.01) both enhanced market performance. Market per-
formance also related to nancial performance ( = 0.75, p < 0.01).
Firm size ( = 0.03, n.s.), as a control variable, had no signicant
effect on performance.
To test for a potential direct relationship between SIC and rm
performance, we considered whether the relationship actually was
mediated by process exibility or cost efciency. According to
Baron and Kenny (1986) and Iacobucci et al. (2007), a variable
serves as a mediator if it can account for the relation between
the predictor and the criterion. We decomposed the co-alignment
model into direct effects, indirect effects and total effects, then
added two direct paths to the co-alignment model. However, the
direct effects of SIC on market performance ( = 0.15, n.s.) and nan-
cial performance ( = 0.05, n.s.) were not signicant. The indirect
paths of SIC on rm performance ranged from 0.045 to 0.051. Fur-
thermore, we used a Sobel (1982) test to assess the mediation
effects formally. The two-tailed z-test (market performance z = 4.98
for process exibility, 4.47 for cost efciency; nancial performance
z = 4.35 for process exibility, 4.71 for cost efciency) was rejected
(at p < 0.01), indicating that both process exibility and cost ef-
ciency served as full mediators of the relationship between SIC and
rm performance. We thus found support for H2.
4.3. Direct effect model
In the test of the proposed co-alignment model (Venkatraman,
1989), we compared it against a direct effect model, in which the
three rst-order capabilities relate directly to process exibility
and cost efciency. That is, we removed the second-order latent
construct (i.e., SIC) from Fig. 1, then added paths from each of
the rst-order integration types to process exibility and cost ef-
ciency. The remaining paths were similar to those in Fig. 2 (see
Appendix A). As proposed by Bollen (1989), we compared the mod-
els according to their overall t statistics, then used component t
statistics to judge the adequacy of individual aspects. For the nal
model selection, we followed the parsimony rules put forth by Hull
et al. (1995). The overall t statistics for the direct effect model were
inferior to those of the second-order model (2 = 787.44, df = 141,
p < 0.01, CFI = 0.872, RMSEA = 0.080, 90% condence interval [0.074,
0.086]) (see Table 3). The chi-square test of the co-alignment model
showed better t with the data; furthermore, the direct effect
model did not reveal signicant direct effects of integration sensing
capability on cost efciency ( = 0.10, n.s.) or of seizing capability
on process exibility ( = 0.09, n.s.), inconsistent with prior liter-
ature. The second-order model also was more parsimonious than
the rst-order model, in that it demanded the estimation of one
fewer path (Hull et al., 1995).
The t statistics and parsimony considerations (Anderson and
Gerbing, 1988; Hull et al., 1995) led us to prefer the second-order
factor model (Fig. 2). In turn, we concluded that complementarities
exist among sensing, seizing and transforming. Furthermore, the
results indicated that rms that possess high SIC can improve their
operational performance, in terms of both process exibility and
cost efciency; such benets did not result from the separate sub-
capabilities of integration sensing and seizing. Firms with SIC thus
can avoid the traditional trade-off between cost and exibility.
4.4. Moderating variables
We conducted multiple-group analyses to examine whether
the effects of SIC on relationship performance were moderated by
external and internal dynamics. First, we ran multi-group anal-
yses for the external (i.e., market and technological) dynamics
by dichotomising the data according to the ve-point market
dynamism (MD) variable: low MD (02) and high MD (35). Sim-
ilarly, we split the data by the level of technological dynamism
(TD): low TD (02) and high TD (35). Second, we ran multi-group
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461 455

Table 5
Measurement invariance.

2 value d.f. NNFI CFI IFI RMSEA 2

Two-group model market dynamics

Congural invariance 536.253 281 0.950 0.959 0.959 0.050
Metric invariance 547.220 293 0.952 0.959 0.959 0.049 10.967
Scalar invariance 543.025 288 0.951 0.959 0.959 0.050 6.772
Structural (path) invariance 559.417 302 0.953 0.959 0.959 0.049 23.164
Two-group model technological dynamics
Congural invariance 378.689 263 0.978 0.983 0.983 0.035
Metric invariance 411.312 281 0.977 0.981 0.981 0.036 32.623
Scalar invariance 519.765 302 0.964 0.968 0.968 0.045 141.076
Structural (path) invariance 525.835 303 0.963 0.967 0.967 0.045 147.146
Two-group model international supply
Congural invariance 432.991 262 0.961 0.970 0.970 0.045
Metric invariance 447.775 274 0.962 0.969 0.970 0.044 14.784
Scalar invariance 592.708 302 0.942 0.948 0.949 0.055 159.717
Structural (path) invariance 546.357 304 0.952 0.957 0.958 0.050 113.366
Two-group model supply complexity
Congural invariance 376.849 262 0.966 0.974 0.974 0.040
Metric invariance 386.647 274 0.968 0.974 0.975 0.038 9.798
Scalar invariance 533.543 295 0.937 0.946 0.946 0.054 156.694
Structural (path) invanriance 471.661 302 0.961 0.921 0.962 0.045 94.812

analyses for the internal dynamics of supply internationalization
(IS) and supply complexity (SC). Because supply international-
ization was measured by the percentage of local suppliers (i.e.,
suppliers within the country/total suppliers), we conducted a
median split of the data points to determine that rms with low
internationalization relied on more than 65% local suppliers. The
supply complexity measure reected the number of key suppliers,
which we obtained by multiplying the total number of suppliers
by the percentage of key suppliers. Again using a median split, we
determined that a manufacturer with a supply base of 20 or more
key suppliers had a highly complex supply base.
In support of the test for external and internal moderators, we
ran individual analyses for each sub-sample to conrm that the
model was valid for each sample. The goodness-of-t and t indices
all exceeded the recommended threshold values (Bollen, 1989;
Byrne, 2006), indicating that the models were acceptable. To test
the moderators, we investigated if the measurement model was
invariant across sub-groups, in which case we would need to test
for differences in the path coefcients that indicated the variable
under analysis was a moderator.
4.4.1. Assessing measurement invariance
The two-group invariance tests featured samples with differ-
ent dynamics (i.e., MD, TD, IS and SC). Previous studies suggest
that in a multi-group approach, the items and number of under-
lying constructs should be the same across groups (Baumgartner
and Steenkamp, 1998; Bollen, 1989); this condition holds in our
research.
Invariance tests should proceed in three steps to determine
congural, measurement and structural invariance (Bollen, 1989;
Byrne, 2006; Steenkamp and Baumgartner, 1998; Vandenberg and
Lance, 2000). Congural invariance tests whether the same num-
ber of factors and factor-loading patterns hold across independent
groups; measurement invariance indicates whether the measure-
ment model is equal across independent groups; and structural
invariance reveals the similarity of the measurement and structural
models across independent groups (Byrne, 2006).
Invariance might be demonstrated by the difference in 2 values
between a restricted model and a baseline congural model. How-
ever, the 2 value is sensitive to sample size and non-normality
(similar to 2 ), so it may not be a practical measure (Byrne, 2006;
Cheung and Rensvold, 2002). Instead, we used two alternative crite-
ria: The multi-group model should demonstrate an adequate t
to the data, and the change in CFI (CFI) values between models
should be negligible (Byrne, 2006; Cheung and Rensvold, 2002).
Invariance tests in an EQS program also make the comparative
approach (2 ) unnecessary, because the Lagrange multiplier (LM)
test provides an immediate identication of non-invariant equality
constraints across groups (Byrne, 2006).
The congural invariance tests for different sample splits began
with a two-group model test of baseline models without any equal-
ity constraints between samples that reveal different dynamics.
As Table 5 shows, the four models exhibit congural invariance;
the model ts the data well. That is, the model structure is opti-
mally represented by the proposed pattern of paths and factor
loadings. Second, we performed a measurement invariance test
to test for metric and scalar. Metric invariance indicates whether
respondents in different groups interpret and respond to measure-
ment items equivalently (Steenkamp and Baumgartner, 1998; Yoo,
2002) by constraining all factor loadings to be equal across groups.
The goodness-of-t results in Table 5 show that the model exhibits
good t with the data. Furthermore, the LM test statistics showed
that all items were invariant across the different two-group sam-
ples, so the interpretation of the measurement items was the same
among the different groups. Scalar invariance refers to consistency
between differences in latent and observed means, which is nec-
essary for comparative validity (Meredith, 1993; Steenkamp and
Baumgartner, 1998; Yoo, 2002). We constrained all invariant factor
loadings and observed variable intercepts, regardless of whether
the factor loading is xed or freely estimated (Byrne, 2006). The
goodness-of-t results again indicated good t between the model
and data, and in the LM test, a limited number of intercepts were
non-invariant. Finally, for the structural invariance testing, we con-
strained the structural paths to be equal across groups and retained
all equality constraints except those parameters that were non-
invariant in the metric invariance test. As we show in Table 5, the
more restricted models t the data well; the LM test results showed
that the structural paths from SIC to process exibility and cost
reduction were non-invariant across some groups.
In Tables 6 and 7, which report the unstandardized path
coefcients, the non-invariant paths are in bold. Thus, in envi-
ronments with high market dynamics, the link between SIC and
operational performance is stronger than in environments with
low market dynamics, in support of H3a. In partial support of H3b,
in dynamic markets, the link of SIC with operational performance
grew stronger through cost efciency, but this result did not hold
for process exibility. The test of the moderating effect of supply
base complexity on operational performance revealed that more
456 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461

Table 6
Structural path model for internal dynamics (MD and TD).

Path coeff. Path coeff. Signicanceof Path coeff. Path coeff. Signicance of
low MD high MD LM univariate low TD high TD LM univariate
n = 216 m = 503 constraint tests n = 280 m = 439 constraint tests

Integration sensing SIC 0.20** 0.19** 0.90 0.20** 0.20** 0.62

Integration seizing SIC 0.26** 0.26** 0.53 0.25** 0.25** 0.67
Integration transforming SIC 0.23** 0.23** 0.44 0.26** 0.26** 0.06
SIC process exibility 0.45** 0.69** <0.01 0.71** 0.71** 0.13
SIC cost reduction 0.47** 0.86** <0.01 0.63** 0.92** <0.01
cost reduction market performance 0.26** 0.26** 0.42 0.22** 0.22** 0.82
cost reduction nancial performance 0.25** 0.25** 0.12 0.22** 0.22** 0.80
process exibility market performance n.s. n.s. 0.10 n.s. n.s. 0.84
process exibility nancial performance n.s. n.s. 0.39 n.s. n.s. 0.95
market performance nancial performance 0.75** 0.75** 0.74 0.74** 0.74** 0.50

Path coefcients in bold show the non-invariant path coefcients when the constraints are released.
We present the unstandardized path coefcients in other to compare the models among the sub-groups.
*
Signicant at p < 0.05.
**
Signicant at p < 0.01.

Table 7
Structural path model for external dynamics (IS and SC).

Path coeff. low Path coeff. high Signicance of Path coeff. low Path coeff. high Signicance of
SC n = 275 SC m = 285 LM univariate IS n = 326 IS m = 326 LM univariate
constraint tests constraint tests

Integration sensing SIC 0.38** 0.38** 0.61 0.20** 0.18** 0.38

Integration seizing SIC 0.45** 0.45** 0.22 0.26** 0.24** 0.27
Integration transforming SIC 0.36** 0.36** 0.76 0.24** 0.25** 0.91
SIC process exibility 0.68** 0.33** <0.01 0.69** 0.70** 0.84
SIC cost reduction 0.81** 0.51** <0.01 0.86** 0.81** 0.86
Cost reduction market performance 0.32** 0.32** 0.69 0.26** 0.22** 0.84
Cost reduction nancial performance 0.28** 0.28** 0.57 0.25** 0.18** 0.06
Process exibility market performance n.s. n.s. 0.98 n.s. n.s. 0.72
Process exibility nancial performance 0.14* 0.14* 0.30 n.s. n.s. 0.08
Market performance nancial performance 0.67** 0.67** 0.57 0.75** 0.73** 0.80

Path coefcients in bold show the non-invariant path coefcients when the constraints are released.
We present the unstandardized path coefcients in other to compare the models among the sub-groups.
*
Signicant at p < 0.05.
**
Signicant at p < 0.01.

key suppliers weakened the relationship, in support of H4a. How-
ever, we found no support for H4b, regarding the impact of supply
internationalization on the link between SIC and operational per-
formance.
4.5. Additional analyses
As is true for all studies that use data from a single source, com-
mon method variance may be a concern, so we conducted two tests.
Harmans one-factor test (Podsakoff and Organ, 1986) indicated
that the model with seven factors showed good t (CFI = 0.972,
RMSEA = 0.0238), whereas that with a single factor showed poor
t (CFI = 0.383, RMSEA = 0.169). This evidence of the presence of
seven different factors indicated that common method effects
were unlikely to have contaminated the results of this data anal-
ysis. However, Podsakoff et al. (2003) critique this test as lacking
sensitivity, so we also followed the procedure recommended by
Widaman (1985) and Podsakoff et al. (2003). The method factor
did not signicantly improve the model and accounted for only a
small portion of variance. All factor loadings of the traits also were
still signicant in the CFA model after we included the common
method factor.
The respondents represented multiple countries, so we next
controlled for cultural effects by splitting the data into four geo-
graphically determined groups: Europe, North America, South
America and Asia. To test for differences across continents, we
investigated whether the general structure of the hypothesized
model differed (Jreskog and Srbom, 1999). In a multi-sample
analysis, with the parameters constrained to be the same across
groups, the LM test did not indicate inappropriateness of our cross-
group constraints at the p < 0.01 level, indicating that the individual
constraints should not be released. Furthermore, the indices for the
constrained model indicated that the data t the different conti-
nents reasonably well (CFI = 0.913, RMSEA = 0.064, 90% condence
interval [0.058, 0.070]). Thus, we found no signicant differences
across these four geographical regions.
Table 8 provides a summary of the hypotheses testing.
5. Discussion
The present research offers an empirical validation of
Teeces (2007) dynamic capability framework in the context of
buyersupplier relationships. Specically, we have dened and
tested the notion of SIC and its ties to operational and nan-
cial performance. Overall, our study recommends a more complex
view of supplier integration than previous studies have offered. It
reveals that complementarities among integration sensing, seiz-
ing and transforming inuence operational performance, in terms
of process exibility (H1a) and cost efciency (H1b) and nally
rm performance (H2). The importance of the integration trans-
forming dimension in linking SIC to performance stresses the need
to regularly revise and adapt supplier integration practices in order
to stay competitive. In addition, recognizing that effective organi-
zational adaptation is partly determined by environmental forces
(Schilke, 2014), our research conrms that the effect of SIC on com-
petitive advantage is stronger when environmental dynamics are
higher (H3) and comparatively lower when the number of sup-
pliers is lower (H4a). We did not nd support for the attenuated
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461 457

Table 8
Summary of ndings.

Path Path coefcients Hypotheses testing

Direct effects
Integration sensing SIC 0.35**
Integration seizing SIC 0.46**
Integration transforming SIC 0.49**
H1a: SIC Process exibility 0.69** H1a supported
H1b: SIC Cost efciency 0.77** H1b supported
Process exibility Market performance 0.27**
Cost efciency Market performance 0.19**
Market performance Financial performance 0.75**
Mediating effects
H2: SIC Process exibility Market performance indirect effect: 0.050; z = 4.98** H2 supported
H2: SIC Cost efciency Market performance indirect effect: 0.045; z = 4.47** H2 supported
H2: SIC Process exibility Financial performance indirect effect: 0.051; z = 4.35** H2 supported
H2: SIC Cost efciency Financial performance indirect effect: 0.050; z = 4.71** H2 supported
Moderating effects
H3a: Market dynamics
SIC Process exibility Low MD: 0.45** ; high MD: 0.69** H3a supported
Difference estimate: 0.24**
SIC Cost exibility Low MD: 0.47** ; high MD: 0.86**
Difference estimate: 0.39**
H3b: Technological dynamics
SIC Process exibility Low TD: 0.71** ; high TD: 0.71** H3b partly supported
Difference estimate: < 0.01
SIC Cost exibility Low TD: 0.63** ; high TD: 0.91** , high TD: 0.92**
Difference estimate: 0.29**
H4a: Supply base complexity
SIC Process exibility Low SC: 0.68** ; high SC: 0.33** H4a supported
Difference estimate: 0.35**
SIC Cost exibility Low SC: 0.81** ; high SC: 0.51**
Difference estimate: 0.30**
H4b: Supply internationalization
SIC Process exibility Low IS: 0.69** ; high IS: 0.70** H4b failed to support
Difference estimate: 0.01
SIC Cost exibility Low IS: 0.86** ; high IS: 0.81**
Difference estimate: 0.05
*
Signicant at p < 0.05.
**
Signicant at p < 0.01.

effect of internationalization of the supply base in linking SIC and
operational performance (H4b).
Different arguments support the complementarity perspective
of integration sensing, seizing and transforming. First, comple-
mentarities result when knowledge gained from one integration
practice enhances the functioning of other integration practices.
Companies that share inventory information (i.e., integration
sensing) will benet more from VMI practices (i.e., integration seiz-
ing), for example. Thus, the presence of one asset makes another
asset more valuable. These sub-capabilities even could be cumu-
lative in nature, though further research is required to test this
prediction. Second, pursuing multiple types of integration simul-
taneously prevents blind spots and can help buyers avoid too
much emphasis on any single set of integration practices, leading
to more balanced investments in resources that support differ-
ent sub-capabilities. Third, integration transforming capabilities,
which reect the continuous effort devoted to new integration
practices, can combine with the current status of integration prac-
tices to enable buyers to evolve and outpace their competition in
designing effective supply chains.
By afrming that SIC positively affects cost efciency and pro-
cess exibility, our analyses also conrm a dynamic view in which
rms take advantage of complementarities to react to changes.
Moreover, SIC can help companies simultaneously strengthen
their process exibility and cost efciency. Using Schmenner and
Swinks (1998) terminology, SIC might help companies break
through their current operating frontier and improve costs and
exibility simultaneously. In particular, SIC enables companies to
remove inefciencies in the supply chain by transforming their cur-
rent processes. These changes do not require massive additional
investments in physical assets but instead can result from changes
to inputs, experiences, planning or controls, which we propose can
be structured as three sub-capabilities: integration sensing, seizing
and transforming.
In line with H2, we did not nd a direct effect of SIC on rm
performance, so it appears that the benets of integration rst
must be translated into operational capabilities (Rosenzweig et al.,
2003). These results align with previous ndings by Ragatz et al.
(1997) that using the knowledge of suppliers to complement inter-
nal capabilities can reduce costs and improve supply chain design.
Moreover, SIC might create competitive offerings for customers,
helping the rm stave off the competition longer, such that it leads
to improved market and nancial performance.
Our conrmation of H3, namely, that market and (partly) tech-
nological dynamics strengthen the relationship between SIC and
operational performance, matches the ndings in supply chain lit-
erature that more dynamic markets and technologies require more
responsive supply chains (Fisher, 1997). However, even with less
dynamic markets and technologies, partners benet from continu-
ously investing in new integration efforts to adapt and improve
their supply chains in the face of new realities. This nding is
in line with strategy literature, in which more dynamic environ-
ments provide more opportunities to cover the cost of developing
a dynamic capability (Zahra et al., 2000).
As Bozarth et al. (2009) argue, we nd that not all sources of
supply chain complexity are disadvantageous. Internal supply base
complexity seems to attenuate the impact of SIC on performance,
but creating an international supply base does not necessarily make
it more difcult to reap the benets of a SIC, perhaps because com-
panies can tailor their supply chains. For example, a company might
458 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
adopt two supply chain congurations: one for high-end/design
products and one for mass products. International supply chain
congurations likely accommodate the mass product solutions,
but local supply chain congurations can create more effective
solutions for design products. Both congurations face challenges
in term of supply chain integration capabilities, but whereas the
local supply chain conguration needs to be fast and exible,
the international supply chain conguration must be reliable and
on-time. Supplier integration capabilities are valuable for both
congurations, but they require different integration practices. Fur-
ther research therefore should examine in more detail the specic
integration practices required with different levels of supply com-
plexity (Gimenez et al., 2012).
5.1. SIC as a tool to manage suppliers and increase performance
Our data afrm that SIC is a higher-order capability that
consists of sensing, seizing and transforming capabilities. These
different capabilities in turn require various skills for their coor-
dinated implementation, involving different people. As suggested
in previous studies, implementing a wider range of supply chain
integration elements is better than implementing a few (Frohlich
and Westbrook, 2001; Rosenzweig et al., 2003; Swink et al., 2007),
though no theory of the complementarities among these different
capabilities has been tested previously. This study identies one
set of complementarities, consisting of integration sensing, seizing
and transforming, derived from dynamic capabilities literature.
By noting these complementarities, our study offers an explana-
tion for why some previous models indicate weak or insignicant
links across operational integration forms and rm performance
(e.g., Rosenzweig et al., 2003; Swink et al., 2007; Vereecke and
Muylle, 2006). Part of the confusion stems from measures of the
impacts of individual integration practices; we considered bundles
of integration practices that enabled rms to not only sense and
seize their supplierbuyer relationships but also transform their
existing practices. These integration practices may not have signif-
icant individual effects and instead may be fully effective only in
the presence of other capabilities. A similar concern arises in the
alternate, direct effect model in Appendix A.
5.2. Theory development about dynamic capabilities
By introducing dynamic capabilities to supplier integration
literature (Teece, 2007), we conceptualize a dynamic nature of
supplier integration. As such, we assert that SIC serves a dual
purpose, operating on both operational and dynamic levels. Pre-
vious research mainly has addressed supplier integration from
an operational perspective, so we stress its dynamic component.
That is, SIC can facilitate smooth delivery processes, as well as
serve as a long-term, knowledge-generating activity to redesign
the supply chain. Various researchers (e.g., Doz and Shuen, 1990;
Mody, 1993) acknowledge that collaborations and partnerships can
lead to organizational learning and help rms recognize dysfunc-
tional practices or prevent strategic blind spots (Teece et al., 1997).
The acquisition and diffusion of inter-organizational knowledge
through such practices may assist in the collection of new infor-
mation that can provide the diversity needed to start building new
knowledge (Zollo and Winter, 1999).
By replicating practices in diverse contexts and accumulating
experience with approaches, organizations simultaneously achieve
multiple objectives. They can complete their adaptation of exist-
ing practices to reect changes in the environments; they also
can renew their existing procedures and apply new practices.
Learning may occur in semi-automatic fashion, on the basis of
the individual adaptations of practices to unsatisfactory perfor-
mance or changes. Teece (2007) describes these adaptations as
transforming processes. Furthermore, Jarrat (2004) describes orga-
nizational learning as a combination of adaptive learning, which
prompts changes consistent with current organizational norms and
behaviour (i.e., seizing), and generative learning, which results in
the applications of new systems, processes and approaches to cre-
ate innovative bases for competitive advantage (i.e., transforming).
Our study stresses the importance of these learning processes to
adapt to new realities.
6. Managerial implications
Firms strive to develop and sustain competitive advantages by
quickly adapting to new business requirements. Our study reveals
that SIC can support this adaptive ability in a cost efcient way. For
example, Walmart has long been known for its supply chain innova-
tions, but even this retail giant requires constant transformations of
its excellent supply chain, because those innovations are constantly
copied and replicated by competitors. Supply chain executives are
thus responsible for constantly investing in new supply chain ini-
tiatives (Johnson, 2008), such as using SIC to collect data about
environmental costs of its products and setting up new green ini-
tiatives together with suppliers (Block, 2013). Along the same line,
the supply chain manager of Sainsbury describes his main respon-
sibility as to develop a culture of change, such that people with
different talents continuously offer ideas to improve the existing
supply chain (Slagmulder et al., 2003). In practice, these changes are
generated through learning from problem-solving exercises and
seeing opportunities for improvement. Moreover, this learning is
created through discussion and debate among managers from the
buyer and supplier side by challenging currently held beliefs on
how to integrate operations among their companies. This require-
ment for continuous adapting to changes creates a new role for
buyers as a project manager instead of a pure negotiator, with
a focus on setting up new supplier integration project to further
improve and synchronize the supply chain.
Firms that have developed integration practices as a coher-
ent set of sensing, seizing and transforming capabilities may be
able to identify, more quickly and intuitively, the gaps in their
integration capabilities. For example, rms with only integration
sensing capabilities may be able to sense what is happening in
the supply chain, but not act on the opportunities, because they
lack the seizing and transforming capabilities. Alternatively, rms
may well possess the capability to strategically recongure the sup-
ply chain (transforming), but may lack the buyersupplier data
(sensing) that are needed to understand the required changes and
to estimate the benets of supply chain recongurations. In other
words, rms that understand which capabilities are needed for sup-
plier integration, will more easily identify and justify the resources
and skills required to build and sustain competitive advantage. A
well-documented example of a company possessing all three SIC
sub-capabilities i.e., integration sensing, seizing and transfor-
ming is ThyssenKrupp Automotive Systems GmbH, a German
company delivering innovative system solutions such as axle
assembly, batteries and tailored tempering to the automotive
industry (Kaufmann et al., 2013). In 2007, ThyssenKrupp Auto-
motive Systems GmbH had launched a supply chain optimization
project, linking the supply process to the customers production
process. As a consequence, the company was admired in the indus-
try for its synchronized supply chain and lean processes. However,
toward the end of 2011, it became clear that further improvements
were necessary to stay ahead of competition. More specically,
erosion of the companys cost-advantage would motivate long-
standing customers to decline follow-up contracts and to start
doing business with competitors. This need to further improve
inspired ThyssenKrupp Automotive Systems GmbH to transform
its supply chain. The existing supply concept, as implemented in
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
2007, consisted of sensing and seizing processes: It contained a
tracking system, which facilitated the exchange of supply chain
data, it provided inventory transparency and it provided data for
calculating the benets of supply chain recongurations (sensing);
It also contained a Just-in-Sequence and sophisticated planning
system, synchronizing the processes of ThyssenKrupp and its sup-
pliers (seizing). In order to sustain its competitive advantage, the
company revised its practices and recongured the existing pro-
cesses (transforming) by launching a new supply chain concept in
2012. An essential element of this new supply chain concept was
the use of an earlier tracking point in the supply chaini.e., track-
ing the supplier capability instead of the arrival time of supply
parts at ThyssenKrupp. This transformation improved the infor-
mation ows, it improved transparency in the supply chain and
it made disruption threats visible sooner. It resulted in a reduction
of safety stock from 4 to 2 days and it guaranteed supply security
two days earlier than before. A cost-benet analysis showed that
this supply chain reconguration helped the company to sustain
its cost advantage relative to its industry competitors (Kaufmann
et al., 2013).
While in todays context, sustainable competitive advantage
or enjoying greater success than current or potential competitors
is rather rare, transient advantage is the new normal (McGrath,
2013). As McGrath (2013, p. 64) argues: In a world where com-
petitive advantage often evaporates in less than a year, companies
cannot afford to spend months at a time crafting a single long-
term strategy. To stay ahead, they need to constantly start new
strategic initiatives, building and exploiting transient competi-
tive advantages. SIC should be seen as a dynamic capability that
enable companies to replace existing resources in their supply
chain and create a better match between the conguration of
a rms resources and changing conditions (Schilke, 2014). As
such, SIC supports companies to focus on longer-term competitive
advantages.
7. Limitations and further research
The chosen focus of our study creates several limitations. We
focused on a limited set of practices related to supplier integration.
Integraon
sensing 1 -0.09*
0.09
Integraon
-0.10
seizing 2
0.35**
0.26**
Integraon 0.31**
transforming 3
* p < 0.05, **p < 0.01
Fig. A1. Direct effect model.
459
Other integration practices also could contribute to improving buy-
ing rms, such as mapping out processes together or joint R&D
projects. Although using a limited set of practices is appropriate for
our choice of reective rather than formative measures, additional
measures could be included in a larger study. We also measured
SIC and its impact on performance from the buyers perspective;
additional research should integrate the suppliers perspective by
collecting dyadic data. We asked buyer respondents to describe
integration practices with key suppliers, so we cannot provide
information about other types of suppliers or differences across
suppliers. Our focus on operational integration and adaption to new
realities excludes other types of integration, such as new prod-
uct development. Nor does our study cover soft characteristics
of collaborative integration such as trust or socialization. In addi-
tion, our data focus on the industrial industry, which might limit
the generalizability of our ndings. This limitation has particu-
lar implications for external dynamics as a moderating variable;
our data offer no information about extreme cases, because by
denition the industrial industry is moderately dynamic. Other
researchers could verify that rms with SIC sustain their per-
formance across environments. The data also come mainly from
European respondents. Although our tests supported the ndings in
other regions, further research could look into possible geograph-
ical differences.
Additional studies could expand the focus of our analysis to
include inter-rm integration, as well as customer integration.
We investigated a dynamic capability using cross-sectional data,
as traditionally done by research into other dynamic capabilities
(e.g., Drnevich and Kriauciunas, 2011; Menguc and Auh, 2006;
Schilke, 2014). With these data, we cannot examine the day-
to-day actions facilitated by each capability. Additional research
might use case studies of individual rms to examine SIC at
work.
Appendix A.
Fig. A1.
0.28**
Process Market
exibility 4 performance 6
0.75**
0.21**
Cost Financial
eciency 5 performance 7
460 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
References
Alfalla-Luque, R., Medina-Lopez, C., Dey, P.K., 2013. Supply chain integration frame-
work using literature review. Prod. Plann. Control: Manage. Oper. 24 (89),
800817.
Amit, R., Schoemaker, P.J.H., 1993. Strategic assets and organizational Rent. Strateg.
Manage. J. 14 (1), 3346.
Anderson, E., Jap, S.D., 2005. The dark side of close relationships. MIT Sloan Manage.
Rev. 46 (3), 7582.
Anderson, J.C., Gerbing, D.W., 1988. Structural equation modeling in practice: a
review and recommended two-step approach. Psychol. Bull. 103 (3), 411423.
Barney, J.B., 1991. Firm resources and sustained competitive advantage. J. Manage.
17 (1), 99120.
Baron, R.M., Kenny, D.A., 1986. The moderatormediator variable distinction in
social psychological research: conceptual, strategic, and statistical consider-
ations. J. Pers. Soc. Psychol. 51 (6), 11731182.
Barratt, M., Oke, A., 2007. Antecedents of supply chain visibility in retail supply
chains: a resource-based theory perspective. J. Oper. Manage. 2 (6), 12171233.
Baumgartner, H., Steenkamp, J.-B., 1998. Multi-group latent variable models for
varying numbers of items and factors with cross-national and longitudinal appli-
cations. Mark. Lett. 9 (1), 2135.
Bentler, P.M., 1992. EQS Structural Equations Program Manual. BMDP Statistical
Software, Inc., Los Angeles, CA.
Bierly, P.E., Coombs, J.E., 2004. Equity alliances, stages of product development, and
alliance instability. J. Eng. Tech. Manage. 21 (3), 191214.
Block, B., 2013. Wal-Mart Scrutinizes Supply-chain Sustainability,
http://www.worldwatch.org/node/6200.
Bollen, K., 1989. Structural Equation with Latent Variables. John Wiley & Sons, New
York, NY.
Boyer, K.K., Lewis, M.W., 2002. Competitive priorities: investigating the need for
trade-offs in operations strategy. Prod. Oper. Manage. 11 (1), 920.
Bozarth, C.C., Warsing, D.P., Flynn, B.B., Flynn, E.J., 2009. The impact of supply com-
plexity on manufacturing plant performance. J. Oper. Manage. 27 (1), 7893.
Braunscheidel, M.J., Suresh, N.C., 2009. The organizational antecedents of a rms
supply chain agility for risk mitigation and response. J. Oper. Manage. 27 (2),
119140.
Byrne, B.M., 2006. Structural equation modeling with EQS: basic concepts, applica-
tions, and programming, second ed. Lawrence Erlbaum Associates, New York,
NY.
Cachon, G.P., Fisher, M., 2000. Supply chain inventory management and the value of
shared information. Manage. Sci. 46 (8), 10321048.
Cai, S., Jun, M., Jang, Z., 2010. Implementing supply chain information integration in
China: the role of institutional forces and trust. J. Oper. Manage. 28 (3), 257268.
Carr, A.S., Pearson, J.N., 1999. Strategically managed buyersupplier relationships
and performance outcomes. J. Oper. Manage. 17, 497517.
Chen, F., Drezner, Z., Ryan, J., Simchi-Levi, D., 2000. Quantifying the Bullwhip effect
in a simple supply chain: the impact of forecasting, lead times, and information.
Inf. Manage. Sci. 46 (3), 436443.
Cheung, G.W., Rensvold, R.B., 2002. Evaluating goodness-of-t indexes for testing
measurement invariance. Struct. Equ. Model. 9, 233255.
Cheung, M.S., Myers, M.B., Mentzer, J.T., 2011. The value of relational learning in
global buyersupplier exchanges: a dyadic perspective and test of pie-sharing
premise. Strateg. Manage. J. 32 (10), 10611082.
Cho, J., Kang, B., 2001. Benets and challenges of global sourcing: perceptions of U.S.
apparel retail rms. Int. Mark. Rev. 18 (5), 542561.
Choi, T.Y., Dooley, K.J., Rungtusanatham, M., 2001. Supply networks and complex
adaptive systems: control versus emergence. J. Oper. Manage. 19 (3), 351366.
Choi, T.Y., Krause, D.R., 2006. The supply base and its complexity: implications for
transaction costs, risks, responsiveness, and innovation. J. Oper. Manage. 24 (5),
637652.
Clark, K.B., Fujimoto, T., 1991. Product development performance. Harvard Univer-
sity Press, Boston, MA.
Claro, D.P., Hagelaar, G., Omta, O., 2003. The determinants of relational governance
and performance: how to manage business relationships? Ind. Mark. Manage.
32 (8), 703716.
Cua, K.O., McKone, K.E., Schroeder, R.G., 2001. Relationships between implementa-
tion of TQM, JIT, and TPM and manufacturing performance. J. Oper. Manage. 19
(6), 675694.
Danneels, E., 2011. Trying to become a different type of company: dynamic capability
at Smith Corona. Strateg. Manage. J. 32 (1), 131.
Deshais, A., Podcast on http://www.ism.ws/content.cfm?ItemNumber=21879 (03
December 2012).
Devaraj, S., Krajewski, L., Wei, J.C., 2007. Impact of eBusiness technologies on opera-
tional performance: the role of production information integration in the supply
chain. J. Oper. Manage. 25 (6), 11991216.
Doz, Y., Shuen, A., 1990. From intent to outcome: a process framework for partner-
ships. In: INSEAD Working Paper. Fontainebleau, France.
Drnevich, P.L., Kriauciunas, A.P., 2011. Clarifying the conditions and limits of the
contributions of ordinary and dynamic capabilities to relative rm performance.
Strateg. Manage. J. 32, 254279.
Droge, C., Vickery, S.K., Jacobs, M.A., 2012. Does supply chain integration mediate
the relationship between product/process strategy and service performance?
An empirical study. Int. J. Prod. Econ. 137 (2), 250262.
Edwards, J.R., 2001. Multidimensional constructs in organizational behavior
research: an integrative analytical framework. Organ. Res. Methods 4 (2),
144192.
Eisenhardt, K.M., Martin, J.A., 2000. Dynamic capabilities: what are they? Strateg.
Manage. J. 21 (1011), 11051121.
Fisher, M.L., 1997. What is the right supply chain for your product? Harv. Bus. Rev.
75, 105116.
Flynn, B.B., Huo, B., Zhao, X., 2010. The impact of supply chain integration on per-
formance: a contingency and conguration approach. J. Oper. Manage. 28 (1),
5871.
Foerstl, K., Hartmann, E., Wynstra, F., Moser, R., 2013. Cross-functional inte-
gration and functional coordination in purchasing and supply management:
antecedents and effects on purchasing and rm performance. Int. J. Oper. Prod.
Manage. 33 (6), 689721.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unob-
servable variables and measurement error. J. Mark. Res. 18, 3950.
Fortune, A., Mitchell, W., 2012. Unpacking rm exit at the rm and the industry
levels: the adaptation and selection of rm capabilities. Strateg. Manage. J. 33
(7), 794819.
Frohlich, M.T., Westbrook, R., 2001. Arcs of integration: an international study of
supply chain strategies. J. Oper. Manage. 19 (2), 185200.
Gimenez, C., van der Vaart, T., van Donk, D.P., 2012. Supply chain integration and per-
formance: the moderating effect of supply complexity. Int. J. Oper. Prod. Manage.
32 (5), 583610.
Govindan, K., 2013. Vendor-managed inventory: a review based on dimensions. Int.
J. Prod. Res. 51 (3), 38083835.
Gunasekaran, A., Ngai, E.W.T., 2004. Information systems in supply chain integration
and management. Eur. J. Oper. Res. 159 (2), 269295.
Harland, C.M., Caldwell, M.D., Powell, P., Zheng, J., 2007. Barriers to supply
chain information integration: SMEs adrift of eLands. J. Oper. Manage. 25 (6),
12341254.
Hatcher, L., 2003. A Step-by-step Approach to Using SAS for Factor Analysis and
Structural Equation Modeling. SAS Institute, Cary, NC.
Helfat, C.E., Finkelstein, S., Mitchell, W., Peteraf, M.A., Singh, H., Teece, D.J., Winter,
S.G., 2007. Dynamic Capabilities: Understanding Strategic Change in Organiza-
tions. Blackwell, Oxford, UK.
Helfat, C.E., Winter, S.G., 2011. Untangling dynamic and operational capabilities:
strategy for the (n)ever-changing world. Strateg. Manage. J. 32 (11), 12431250.
Hu, L.T., Bentler, P., 1995. Evaluating model t. In: Hoyle, R.H. (Ed.), Structural Equa-
tion Modeling, Concepts, Issues, and Applications. Sage Publications, London.
Hull, J.G., Tedlie, J.C., Lehn, D.A., 1995. Modeling the relation of personality variables
to symptom complaints: the unique role of negative affectivity. In: Hoyle, R.H.
(Ed.), Structural Equation Modeling: Issues and Applications. Sage Publications,
Newbury Park, CA.
Iacobucci, D., Saldanha, N., Deng, X., 2007. A meditation on mediation: evidence
that structural equations models perform better than regressions. J. Consum.
Psychol. 17 (2), 140154.
Im, G., Rai, A., 2008. Knowledge sharing ambidexterity in long-term interorganiza-
tional relationships. Manage. Sci. 54 (7), 12811296.
Jacobs, M.A., Vickery, S., Droge, C., 2007. The effects of product modularity on com-
petitive performance: do integration strategies mediate the relationship? Int. J.
Oper. Prod. Manage. 27, 10461068.
Jarrat, D., 2004. Conceptualizing a relationship management capability. Mark. The-
ory 4 (4), 287309.
Johnson, F., 2008. Supply Chain Management at Wal-Mart. Ecch Case Study
9B07D001. Richard Ivey School of Business, Canada, pp. 115.
Jreskog, K.G., Srbom, D., 1999. LISREL 8.3: User Reference Guide. Scientic Soft-
ware International, Chicago, IL.
Kale, J., Dyer, J.H., Singh, H., 2002. Alliance capability, stock market response, and
long-term alliance success: the role of the alliance. Strateg. Manage. J. 23 (8),
747767.
Katunzi, T.M., 2011. Obstacles to process integration along the supply chain: man-
ufacturing rms perspective. Int. J. Bus. Manage. 6, 105113.
Kaufmann, L., Reimann, F., Khl, T., Lohmann, B., Reinholz, F., Schrder, T., 2013. Case
ThyssenKrupptackling the supply chain challenge: from best practice perfor-
mance to supply chain excellence. In: Case from the Case Centre. ThyssenKrupp
Academy, pp. 30 pp (reference no. 613-030-1). www.thecasecentre.org.
Kim, K.K., Ryoo, S.Y., Jung, M.D., 2011. Inter-organizational information systems
visibility in buyersupplier relationships: the case of telecommunication equip-
ment component manufacturing industry. Omega 39 (6), 667676.
Kotha, S., Orne, D., 1989. General manufacturing strategies: a conceptual synthesis.
Strateg. Manage. J. 10 (3), 211231.
Krause, D.R., Handeld, R.B., Tyler, B.B., 2007. The relationships between sup-
plier development, commitment, social capital accumulation and performance
improvement. J. Oper. Manage. 25 (2), 528545.
Kulp, S.C., Lee, H.L., Ofek, E., 2004. Manufacturer benets from information integra-
tion with retail customers. Manage. Sci. 50 (4), 431444.
Lage, M.J., Filho, M.G., 2010. Variations of the Kanban system: literature review and
classication. Int. J. Prod. Econ. 125 (1), 1325.
Lee, H.L., So, K.C., Tang, C.S., 2000. The value of information sharing in a two-level
supply chain. Manage. Sci. 46 (5), 626643.
Lindberg, P., Voss, C.A., Blackmon, K., 1998. International Manufacturing Strategies,
Context, Content, and Change. Kluwer Academic Publishers, Dordrecht, NL.
McGrath, R.G., 2013. Transient advantage. Harv. Bus. Rev. 91 (6), 6270.
McKone, K., Schroeder, R., Cua, K., 1999. Total productive maintenance: a contextual
view. J. Oper. Manage. 17 (2), 123144.
Menguc, B., Auh, S., 2006. Creating a rm-level dynamic capability through cap-
italizing on market orientation and innovativeness. J. Acad. Mark. Sci. 34 (1),
6373.
 E. Vanpoucke et al. / Journal of Operations Management 32 (2014) 446461
Menor, L.J., Kristal, M.M., Rosenzweig, E.D., 2007. Examining the inuence of oper-
ational intellectual capital on capabilities and performance. Manuf. Serv. Oper.
Manage. 9 (4), 559578.
Meredith, W., 1993. Measurement invariance, factor analysis, and factorial invari-
ance. Psychometrika 58 (4), 525543.
Milgrom, P., Roberts, J., 1995. Complementarities and t strategy, structure and
organizational change in manufacturing. J. Acc. Econ. 19 (23), 179208.
Miller, J.G., Roth, A.V., 1994. A taxonomy of manufacturing strategies. Manage. Sci.
40 (3), 285304.
Mody, A., 1993. Learning through alliances. J. Econ. Behav. Organ. 20 (2), 151170.
Mohr, J., Spekman, R., 1994. Characteristics of partnership success: partnership
attributes, communication behavior and conict resolution techniques. Strateg.
Manage. J. 15, 135152.
Monczka, R.M., Petersen, K.J., Handeld, R.B., Ragatz, G.L., 1998. Success factors in
strategic supplier alliances: the buying company perspective. Decis. Sci. 29 (3),
553577.
Mller, M., Gaudig, S., 2011. An empirical investigation of antecedents to information
exchange in supply chains. Int. J. Prod. Res. 49 (6), 15311555.
Narasimhan, R., Kim, S.W., 2002. Effect of supply chain integration on the rela-
tionship between diversication and performance evidence from Japanese and
Korean rms. J. Oper. Manage. 20, 303323.
Narasimhan, R., Kim, S.W., Tan, K.C., 2008. An empirical investigation of supply chain
strategy typologies and relationships to performance. Int. J. Prod. Res. 46 (18),
52315259.
OLeary-Kelly, S.W., Vokurka, R.J., 1998. The empirical assessment of construct valid-
ity. J. Oper. Manage. 16 (4), 387405.
Peng, D.X., Schroeder, R.G., Shah, R., 2008. Linking routines to operations capabilities:
a new perspective. J. Oper. Manage. 26 (6), 730748.
Penrose, E.T., 1959. The Theory of the Growth of the Firm. Wiley, New York, NY.
Peteraf, M.A., 1993. The cornerstones of competitive advantage. Strateg. Manage. J.
14 (3), 179191.
Peteraf, M., Di Stefano, G., Verona, G., 2013. The elephant in the room of dynamic
capabilities: bringing two divergent conversations together. Strateg. Manage. J.,
http://dx.doi.org/10.1002/smj.2078.
Pierce, L., Boerner, C., Teece, D.J., 2002. Dynamic capabilities, competence, and the
behavioral theory of the rm. In: Augier, M., March, J.G. (Eds.), The Economics of
Choice, Change and Organization: Essays in Honor of Richard M. Cyert. Edward
Elgar, Cheltenham, UK, pp. 8195.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P., 2003. Common method
biases in behavioural research: a critical review of the literature and recom-
mended remedies. J. Appl. Psychol. 88 (5), 879903.
Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research: problems
and prospects. J. Manage. 12, 531544.
Protogerou, A., Caloghirou, Y., Lioukas, S., 2011. Dynamic capabilities and their indi-
rect impact on rm performance. Ind. Corp. Change 21 (3), 615647.
Ragatz, G.L., Handeld, R.B., Petersen, K.J., 2002. Benets associated with supplier
integration into new product development under conditions of technology
uncertainty. J. Bus. Res. 55 (5), 389400.
Ragatz, G., Handeld, R., Scannell, T., 1997. Success factors for integrating suppliers
into new product development. J. Prod. Innov. Manage. 14 (3), 190202.
Rodrigues, A.M., Stank, T.P., Lynch, D.F., 2004. Linking strategy, structure, process,
and performance in integrated logistics. J. Bus. Logist. 25, 6594.
Rosenzweig, E.D., Roth, A.V., Dean Jr., J.W., 2003. The inuence of an integration
strategy on competitive capabilities and business performance: an exploratory
study of consumer products manufacturers. J. Oper. Manage. 21 (4), 437456.
Roth, A.V., Gif, C., Seal, G., 1992. Operating strategies for the 1990s: elements com-
prising world-class manufacturing. In: Voss, C. (Ed.), Manufacturing Strategy,
Process and Content. Chapman & Hall/Norstrand Publishers, London, UK, pp.
133165.
Sazadeh, M.H., Ritzman, L.P., Mallick, D., 2000. Revisiting alternative theoretical
paradigms in manufacturing strategy. Prod. Oper. Manage. 9 (2), 111127.
Sanchez, R., Heene, A., 1996. Competence-based strategic management: concepts
and issues for theory, research and practice. In: Heene, A., Sanchez, R. (Eds.),
Competence Based Strategic Management. John Wiley and Sons, Chichester.
Savitz, E., 2012. Forbes.com, http://www.forbes.com/sites/ciocentral/2012/10/04/
managing-the-risks-of-a-globalized-supply-chain/.
Schilke, O., 2014. On the contingent value of dynamic capabilities for competi-
tive advantage: the nonlinear moderating effect of environmental dynamism.
Strateg. Manage. J. 35 (2), 179203.
Schmenner, R.W., Swink, M.L., 1998. On theory in operations management. J. Oper.
Manage. 17, 97113.
461
Schumpeter, J.A., 1950. Capitalism, Socialism and Democracy. Harper, New York, NY.
Shah, R., Goldstein, S.M., 2006. Use of structural equation modeling in operations
management research. J. Oper. Manage. 24 (2), 148169.
Sila, I., 2007. Examining the effects of contextual factors on TQM and performance
through the lens of organizational theories: an empirical study. J. Oper. Manage.
25 (1), 83109.
Simonin, B.L., 1999. Ambiguity and the process of knowledge transfer in strategic
alliances. Strateg. Manage. J. 20 (7), 595623.
Skinner, W., 1974. The decline, fall and renewal of manufacturing. Ind. Eng. 6 (10),
3238.
Slagmulder, R., Grottoli, D., Corsten, D., 2003. Sainsburys (A): transforming the sup-
ply chain. In: Ecch Case Study 603-020-1. INSEAD and University of St. Gallen,
pp. 113.
Smaros, J., 2007. Forecasting collaboration in the European grocery sector: observa-
tions from a case study. J. Oper. Manage. 25 (3), 702716.
Sobel, M.E., 1982. Asymptotic intervals for indirect effects in structural equa-
tions models. In: Leinhart, S. (Ed.), Sociological Methodology. Jossey-Bass, San
Francisco, CA, pp. 290312.
Sousa, R., Voss, C., 2008. Contingency research in operations management practices.
J. Oper. Manage. 26 (6), 697713.
Steenkamp, J.-B.E.M., Baumgartner, H., 1998. Assessing measurement invariance in
cross-national consumer research. J. Consum. Res. 25 (1), 7890.
Swink, M., Narasimhan, R., Wang, C., 2007. Managing beyond the factory walls:
effects of four types of integration on manufacturing plant performance. J. Oper.
Manage. 25, 148164.
Teece, D.J., 2007. Explicating dynamic capabilities: the nature and microfoundations
of (sustainable) enterprise performance. Strateg. Manage. J. 28, 13191350.
Teece, D.J., Pisano, G., Shuen, A., 1997. Dynamic capabilities and strategic manage-
ment. Strateg. Manage. J. 18 (7), 509533.
Underwood, A., Agg, S., 2012. Supply Chain Agility: Managing Change. A Study of
Supply Chain Maturity November 2012. KPMG International, London, U.K., pp.
132.
Vandenberg, R.J., Lance, C.E., 2000. A review and synthesis of the measurement
invariance literature: suggestions, practices, and recommendations for organi-
zational research. Organ. Res. Methods 3 (1), 470.
Van de Vijver, F.J.R., Leung, K., 1997. Methods and Data Analysis for Cross-cultural
Research. Sage Publications, Inc., Thousand Oaks, California.
Vanpoucke, E., Boyer, K.K., Vereecke, A., 2009. Supply chain information ow strate-
gies: an empirical taxonomy. Int. J. Oper. Prod. Manage. 29 (12), 12131241.
Venkatraman, N., 1989. The concept of t in strategy research: toward verbal and
statistical correspondence. Acad. Manage. Rev. 14 (3), 423444.
Vereecke, A., Muylle, S., 2006. Performance improvement through supply chain col-
laboration in Europe. Int. J. Oper. Prod. Manage. 26 (11), 11761198.
Villena, V.H., Revilla, E., Choi, T.Y., 2011. The dark side of buyersupplier relation-
ships: a social capital perspective. J. Oper. Manage. 29 (6), 561576.
Vollmann, T., Berry, W., Whybark, D.C., Jacobs, R., 2005. Manufacturing planning and
control for supply chain management, fth ed. McGraw-Hill/Irwin, New York,
NY.
Wagner, S.M., Krause, D.R., 2009. Supplier development: communication
approaches, activities and goals. Int. J. Prod. Res. 47 (12), 31613177.
Webster, 1966, 1966. Westers Third New International Dictionary. William Benton,
Chicago, IL.
Widaman, K.F., 1985. Hierarchically nested covariance structure models for research
multitrait-multimethod data. Appl. Psychol. Meas. 9, 126.
Wiengarten, F., Pagell, M., Ahmed, M.U., Gimeneza, C., 2014. Do a countrys logisti-
cal capabilities moderate the external integration performance relationship? J.
Oper. Manage. 32 (1), 5163.
Winter, S.G., 2003. Understanding dynamic capabilities. Strateg. Manage. J. 24,
991995.
Yoo, B., 2002. Cross-group comparisons: a cautionary note. Psychol. Mark. 19 (4),
357368.
Zahra, S.A., Ireland, R.D., Hitt, M.A., 2000. International expansion by new venture
rms: international diversity, mode of market entry, technological learning, and
performance. Acad. Manage. J. 43 (5), 925950.
Zhao, X., Huo, B., Flynn, B.B., Yeung, J.H.Y., 2008. The impact of power and relation-
ship commitment on the integration between manufacturers and customers in
a supply chain. J. Oper. Manage. 26 (3), 368388.
Zhou, H., Benton, W.C., 2007. Supply chain practice and information sharing. J. Oper.
Manage. 25 (6), 13481365.
Zollo, M., Winter, S.G., 1999. From organizational routines to dynamic capabilities.
In: INSEAD Working Paper. Fontainebleau, France.