Int. J.

Production Economics 133 (2011) 489495

Contents lists available at ScienceDirect

Int. J. Production Economics

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

On the complementarity between internal and external just-in-time bundles to build and sustain high performance manufacturing
Andrea Furlan a,n, Giorgia Dal Pont b, Andrea Vinelli b
a b

Department of Economics, University of Padova, Via Del Santo, 33, 35100 Padova, Italy Department of Management and Engineering, University of Padova, Vicenza, Italy

a r t i c l e in fo
Article history: Received 4 June 2009 Accepted 29 July 2010 Available online 7 August 2010 Keywords: Internal just-in-time Upstream just-in-time Downstream just-in-time Operational performance Complementarity

abstract
The focus on the sequential nature of the improvement initiatives has neglected the synergistic effects among practices and literature lacks research on complementarity between internal and external bundles of practices. The aim of this paper is to test the existence of complementarity among internal and external just-in-time bundles. We run statistical analysis using the third round of High Performance Manufacturing international research project data set and we nd that upstream and downstream JIT are complements. This nding suggests the importance of managing the interdependencies both in designing and implementing upstream and downstream JIT in order to maximize operational performance. & 2010 Elsevier B.V. All rights reserved.

1. Introduction Research and practice have clearly shown how internal and external integrations lead to overall improvements in operational performances, customer service, relationships among supply network members, levels of inventory and lead times (Cooper et al., 1997; Harland et al., 1999; Simchi-Levi et al., 2003). However, achieving integration is not an easy task and it cannot occur overnight. Each company involved in integration initiatives should rst enhance internal integration and then, gradually, embrace inter-organizational integration within the supply network (Cooper and Ellram, 1993). Several studies have indeed emphasized that the rationalization and improvement of internal processes must precede practices of integration along the supply chain (Scott and Westbrook, 1991; Olsen and Ellram, 1997). These recognitions notwithstanding, Operations Management (OM) literature has not delved into the tangled relationships between internal integration practices, external integration practices and operational performances. In particular, in line with recent contributions (Siggelkow, 2001; Cassiman and Veugelers, 2006), we maintain that OM literature has not adequately searched for complementarity between internal and external integration practices, i.e. to what extent internal integration practices increase the marginal returns of external integration practices, and vice-versa. This is a nontrivial issue since the

Corresponding author. Tel.: +39 049 8274235; fax: + 39 049 8274211. E-mail addresses: andrea.furlan@unipd.it (A. Furlan), giorgia.dalpont@unipd.it (G. Dal Pont), andrea.vinelli@unipd.it (A. Vinelli). 0925-5273/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2010.07.043

existence of complementarity has in fact relevant implications both in designing and implementing integration initiatives (Milgrom and Roberts, 1995; Siggelkow, 2001; Siggelkow, 2002a, 2002b; Cassiman and Veugelers, 2006). In this paper we focus on the implementation programs of just-in-time (JIT) practices along the supply chain. These programs are aiming at integrating the logistic ows both within the focal rm and between the focal rm and upstream suppliers and downstream customers (Ballou, 1992; Hahn et al., 1983; Dong et al., 2001). Indeed, JIT includes those practices related to stream production ow, such as lot size reduction, cycle time reduction, quick changeover techniques and production process reengineering (Shah and Ward, 2003). JIT is one of the main facet of lean manufacturing, i.e. a managerial philosophy and a set of integrated socio-technical practices aimed at eliminating waste along the whole value chain within and across companies (Womack et al., 1990; Holweg, 2007). On a managerial level this approach turns into a collection of sets of practices and techniques that both implement and support lean philosophy. JIT is one of these sets of practices (Dal Pont et al., 2008). The aim of this paper is to study the complementarity among internal JIT, upstream (towards suppliers) JIT and downstream (towards customers) JIT. We also test the complementarity between downstream JIT and upstream JIT. According to previous literature (Shah and Ward, 2003; Slack et al., 2007), we consider the classic operational performance areas: quality, dependability, exibility and cost. The paper is organized as follows. The second section reviews the relevant literature on just-in-time as one of the most relevant

490

A. Furlan et al. / Int. J. Production Economics 133 (2011) 489495

lean manufacturing bundle. We highlight the interdependencies between internal and external JIT practices and rely on the theory of complementarity to advance a set of testable hypothesis. The third section provides the methodological background and research design. The fourth section presents statistical results and analyses while the fth section discusses theoretical and managerial implications. The nal section highlights the limitations of our study and outlines avenues for future research.

the rm. Buyers can reap benets on quality, cost and deliveries by engaging suppliers in adopting JIT practices in the manufacturing and distribution activities. Simchi-Levi et al. (2003) maintain that strategic partnership between suppliers and manufacturers may have a signicant impact on supply chain performance. Kulp et al. (2004) nd a strong link between an effective customer supplier logistics coordination and operational performance. Stank et al. (2001) and Gimenez and Ventura (2005) demonstrate that a higher level of external logistics integration leads to higher performance.

2. Literature review 2.1. The importance of internal and external integration: JIT as an integrated system One of the challenges that rms have to cope with nowadays is the search for high levels of operational performance through the adoption of an integrated and coherent system of innovative practices (Womack et al., 1990; Schroeder and Flynn, 2001). Several studies have shown how integration often leads to competitive advantage or improvement in the operational performances (Ettlie and Stoll, 1990; Sazadeh et al., 1996; Pagell et al., 2000). Organizational literature has mainly focused on the integration between dyads of internal functions. Hill (1994) and Verma et al. (2001) show how a close collaboration between purchasing and marketing leads to enhanced level of protability of the rm. Narasimhan and Das (2001) demonstrate the positive correlation between the integration of the purchasing function and its practices and the operational performances. Similarly, Operations Management literature highlights the importance of integration among different practices (or bundles of practices) within the rm. Several papers show that integration between different sets of practices leads to the maximization of the operational performance. For example, although Total Quality Management (TQM) and JIT work separately, their combination leads to further performance improvements (Sriparavatsu and Gupta, 1997; Mefford, 1989; Flynn et al., 1999). TQM practices improve JIT performance by reducing manufacturing process variance that leads to safety stock inventory reduction and yields shorter cycle times. On the other hand, JIT practices affect quality by reducing potential scrap and reworking resulting from process failure. OM Scholars have also considered the need for external integration (Watts et al., 1992). Several scholars argue that companies no longer compete as solely players but as members of an integrated supply chain. Achieving integration along the whole supply chain becomes a key success factor to compete in todays ever complex and competitive markets (Frohlich and Westbrook, 2001). Tan et al. (1998) demonstrate that when companies integrate and act as a single entity, performance is enhanced throughout the supply chain while Lee and Billington (1992) argue that integrating suppliers and customers could spark new opportunities for improvement at the supply chain level. In their seminal work, Frohlich and Westbrook (2001) present how that rms extensively integrating both upwards (i.e. towards suppliers) and downwards (i.e. towards customers) reap higher performances than rms adopting other integration strategies. Within this last stream of research scholars have focused primary on supply chain logistics integration, in particular showing that integration between internal and external just-intime practices leads to the maximization of operational performance. Just-in-time initiatives, aimed at streamlining the supply chain, improve the coordination between the supply chain partners and enhance supply chain performance (Lamming, 1993). Helper (1991) shows how extending JIT practices to suppliers is crucial to improve the long-term competitiveness of 2.2. Internal and external just-in-time practices: implementation process and complementarity The implementation of internal and external just-in-time practices is not an easy process to carry out. Even though it is not enough to determine the success of an integration process, many scholars maintain that the process itself should follow a time sequence that starts with the internal integration and ends with the external one. Stevens (1989) claims that the integration process should follow four stages: no integration, internal integration, company integration and nally external integration. Cooper and Ellram (1993) share this point of view. They maintain that rms should improve their internal integration before getting involved in external integration with the other members of the supply chain. Along this vein, Scott and Westbrook (1991) point out that the rationalization and the improvement of the internal activities should precede the integration practices with suppliers and customers. Bowersox (1989) focuses on internal logistical process and suggests that supply chain integration process should start from the internal integration of the operational and logistical processes and then involve customers and suppliers. Cagliano et al. (2006) underline the fact that the lack of internal integration is one of the reasons why projects of external integration fail. Since the adoption of internal integration inuences external integration, managers should adopt a holistic approach to the integration initiatives. An example is the use of Kanban as a method of just-in-time to implement a pull control along the supply network. First a rm should adopt Kanban to trigger the movement of materials within its production process to get improvement in terms of fall-off in overproduction and work in progress. When internal JIT is successfully in place and thoroughly operational, the rm should start implementing JIT towards customers and suppliers using vendor-Kanban for raw materials and outsourced components. The implementation process is not only a matter of a sequence of predetermined stages. It is argued that internal and external integration practices are in fact interwoven and support each other over time. Some scholars maintain that the implementation of just-in-time practices, both internally and externally oriented, improve the performances of the rm because lean practices, of which just-in-time is their epitome, though separated, are highly interrelated (Womack and Jones, 1996; Flynn et al., 1999; Shah and Ward, 2007). Das et al. (2006) highlight how internal logistics integration acts as a catalyst in the development and implementation of the external integration practices because it creates a strategic linkage between these practices and the competitive aims of the rm. Rosenzweig et al. (2003) claim that internal integration is a rst step towards the whole supply chain integration, but, at the same time, integrating with suppliers and customers can spur opportunities for internal integration. Stank et al. (2001) nd that to improve its performance the rm has to reach high external integration through an increase of

A. Furlan et al. / Int. J. Production Economics 133 (2011) 489495

491

internal integration. On the other hand, internal integration can benet from feedback by suppliers and/or customers. Others reach similar conclusions. Gimenez and Ventura (2005) demonstrate that internal and external integrations are positively correlated. Internal integration facilitates the coordination between the various members of the supply chain and external integration supports alignment between different parts of the organization. On the whole, these studies show that it is important to analyze the interdependencies between internal and external just-in-time practices to gain a better understanding of how rms can use these practices to achieve better performances. We rely on the theory of complementarity to reach this objective. The theory of complementarity provides the basis to understand how various elements of strategy, structure and processes of an organization are interrelated (Milgrom and Roberts, 1990, 1995). The concept of complementarity offers a rigorous explanation to the synergistic effects among activities. Two activities are complementary when the adoption of one increases the marginal returns of the other and vice-versa (Cassiman and Veugelers, 2006). Knowing if two activities (or sets of activities) are complements has important implications for their deployment and implementation processes. The study of complementarity between activities can be traced back to the theory of supermodularity (Milgrom and Roberts, 1990, 1995). This theory can be used in research seeking to empirically validate the existence of complementarity (MacDufe, 1995; Cassiman and Veugelers, 2006; Miravete and Pernias, 2006). Suppose that a rm has two activities, A1 and A2, and each of them could be performed by the rm (Ai1) or not (Ai0), i{1, 2}. The associated performance function P(A1, A2) is supermodular and A1 and A2 are complements only if

marginal returns of upstream JIT on operational performance and vice-versa. Hypothesis 2. Internal JIT bundle and downstream JIT bundle are complementary: the implementation of internal JIT increases the marginal returns of downstream JIT on operational performance and vice-versa. Hypothesis 3. Upstream JIT bundle and downstream JIT bundle are complementary: the implementation of upstream JIT increases the marginal returns of downstream JIT on operational performance and vice-versa. 3. Methodology and research design 3.1. Sample The data used for empirical analysis are part of the third round of the High Performance Manufacturing (HPM) international project. Data were collected from 2005 to 2007 on 266 manufacturing plants located in nine countries (Finland, Sweden, Germany, Japan, Korea, Austria, Italy, Spain and the United States) and three industries (electronics, machinery and transportation). The sample distribution by country and industry is provided in Table 1. The HPM data set provides a stratied sample consisting of traditional and world-class manufacturing plants. World-class manufacturers were randomly selected from a list of plants that won one or more industry awards and/or had been signaled by experts opinions published in trade journals and industrial publications. The traditional manufacturers were randomly selected from industrial publications such as Duns and Kompass. Plants larger than 100 employees were selected. Plants managers were contacted via phone calls to check for the rm availability to participate in the project. The questionnaires were then sent to the plant coordinator, who distributed and collected the questionnaire from 10 managers (i.e. plant accounting manager, human resource manager, information system manager, production control manager, inventory manager, product design engineer, process engineer, plant manager, quality manager and plant superintendent), 2 supervisors and direct workers in each plant. The third round of the HPM project yielded a response rate of 65% (calculated as the percentage of plants contacted by the research team that returned the surveys). 3.2. Variables and measures Relying on previous relevant literature (Shah and Ward, 2003, 2007; McLachlin, 1997; Flynn et al., 1999; Dal Pont et al., 2008),
Table 1 Sample industry and country distribution. Industry Electronics Country Finland Germany Italy Japan South Korea Spain Sweden United States Austria Total 14 9 10 10 10 9 7 9 10 88 Machinery 6 13 10 12 10 9 10 11 7 88 Transportation 10 19 7 13 11 10 7 9 4 90 30 41 27 35 31 28 24 29 21 266 Total

P1, 1-P0, 1 Z P1, 0-P0, 0

Complementarity exists between the two activities if

y11 -y01 Z y10 -y00

where yjk (j {0, 1} and k {0,1}) are the marginal returns on performance P associated to activity 1 and activity 2. In other words, adding an activity while the other is already being performed has a higher incremental effect on performance (P) than adding the activity in isolation. By the same token, an activity is less benecial if the other has already been abandoned (Bocquet et al., 2007). The complementarity between bundles of managerial practices implies that the implementation of one bundle has a positive effect on the implementation of the others (Athey and Stern, 1998). Even if the concept of complementarity is increasingly used in management literature, effective empirical applications are scant, due to the fact that the concept often remains illdened (Cassiman and Veugelers, 2006). This paper seeks to empirically test the complementarity among internal and external JIT practices. In particular, we study the complementarity among internal JIT, upstream JIT (i.e. towards suppliers) and downstream JIT (i.e. towards customers). On the basis of the previously mentioned OM literature that highlights the interdependencies between internal and external JIT practices and using the theory of complementarity that offers a rigorous tool to test the synergistic effect among activities, we advance the following hypotheses: Hypothesis 1. Internal JIT bundle and upstream JIT bundle are complementary: the implementation of internal JIT increases the

492

A. Furlan et al. / Int. J. Production Economics 133 (2011) 489495

we select 5 items to measure internal JIT, 7 items for upstream JIT and 6 items for downstream JIT. We control our results for the age of the plant (measured as the number of years since inception) and the size of the plant (measured as the number of employees). The items concerning JIT both internal and external are measured on a 17 Likert scale. Informants are asked for their perception of the actual implementation of each practice. The items concerning operational performance are measured on a 15 Likert scale, and informants are asked for their perception of realized performance. In line with existent literature, we treat operational performance as a single latent construct (Shah and Ward, 2003, 2007). The use of perceptive measures allows researchers to gather data from different respondents on a wide range of topics. The common method bias was controlled originally in the data set including reverse coding questions. Content validity of the measurement items is assessed by including items that have been used in prior studies. We also run CFA to provide further evidence of content validity. Hair et al. (1998) suggest 0.30 as the minimum acceptable threshold for factor loadings. For all the constructs, every item is statistically signicant and exceeds this threshold level. Moreover, internal reliability was tested through the Cronbachs a index, which for all the constructs is higher than the recommended threshold of 0.6.

Table 2 shows the statements used to measure each lean bundle and operational performance and provides the results of reliability and validity analysis.

4. Analysis and results To test the existence of complementarity we follow the procedure of Cassiman and Veugelers (2006). First of all we create dummy variables that indicate whether the rm performs internal and external JIT. We take the median of the distribution as the threshold to create the dummies. A more generous approach would have been to take the 1st and 4th quartiles, but the adoption of the median allows us to keep as much information as possible. For internal JIT the median is 4.733, for upstream JIT the median is 4.143 and for downstream JIT the median is 4.757. We run t-test for mean differences to detect if these thresholds statistically discriminate each bundle. All the tests indicate that the medians discriminate each bundle [internal JIT t-test: 21.452 (p 0.000); upstream JIT t-test: 22.610 (p0.000); downstream JIT the t-test: 22.538 (p 0.000)]. Using these dummy variables, we construct twelve exclusive categories. For example, to test Hypothesis 1, i.e. the complementarity between internal JIT and upstream JIT, we construct four groups: (1) rms that have a low degree in internal JIT and upstream JIT (Low IntJIT&UpJIT); (2) rms that have high degree of internal JIT but a low degree of upstream JIT (Mainly IntJIT); (3) rms that have low degree of internal JIT but a high degree of upstream JIT (Mainly UpJIT); and (4) rms that have high degree of both internal and upstream JIT (High IntJIT&UpJIT). Similarly, we construct four groups to test Hypothesis 2, i.e. the complementarity between internal JIT and downstream JIT (i.e. Low IntJIT&DownJIT, Mainly IntJIT, Mainly DownJIT and High IntJIT&DownJIT) and four groups to test Hypothesis 3, i.e. the complementarity between upstream and downstream JIT (Low UpJIT&DownJIT, Mainly UpJIT, Mainly DownJIT and High UpJIT&DownJIT). Table 3 reports the descriptive statistics of the constructs involved in the study. We test the existence of complementarity with the direct approach suggested by Cassiman and Veugelers (2006). We perform three OLS regressions to test the complementarity between internal JIT and upstream JIT (Table 4), internal JIT and downstream JIT (Table 5), and upstream JIT and downstream JIT (Table 6). We control our results for age since inception and size of the plant. Our results conrm Hypothesis 3 (Table 6) while they reject Hypotheses 1 and 2 (Tables 4 and 5). In particular, regression results in Table 6 show that coefcients of High UpJIT&DownJIT Mainly UpJIT, Mainly DownJIT and
Table 3 Correlation matrix. Mean S.D. IntJIT UpJIT DownJIT Oper. perf. Age

Table 2 Variable and measures. Bundle Internal JIT (a 0.673) Item JSFTN06 Description

We usually complete our daily schedule as planned. JSMHN06 The layout of our shop oor facilitates low inventories and fast throughput. JSPLN06 We use a Kanban pull system for production control. JSSUN04 We have low setup times of equipment in our plant. JSSUN11 We emphasize small lot sizes, to increase manufacturing exibility. JSVNN01 JSVNN02 JSVNN03 JSVNN04 Our suppliers deliver to us on a just-in-time basis. We receive daily shipments from most suppliers. Suppliers ll our Kanban containers, rather than lling purchase orders. Our suppliers deliver to us in Kanban containers, without the use of separate packaging. We can depend upon on-time delivery from our suppliers. Our suppliers are linked with us by a pull system. Suppliers frequently deliver materials to us. Our customers receive just-in-time deliveries from us. Most of our customers receive frequent shipments from us. We always deliver on time to our customers. We can adapt our production schedule to sudden production stoppages by our customers. Our customers have a pull type link with us. Our customers are linked with us via JIT systems. Unit cost of manufacturing. Conformance to product specications. On-time delivery performance . Flexibility to change volume.

Upstream JIT (a 0.818)

JSVNN09 JSVNN10 JSVNN11 Downstream JIT (a 0.757) JSVCN01 JSVCN02 JSVCN04 JSVCN05

JSVCN06 JSVCN07 Operational performance (a 0.630) GRCPN01 GRCPN02 GRCPN03 GRCPN06

IntJIT 4.733 0.691 1 UpJIT 4.170 0.796 0.636nn 1 DownJIT 4.706 0.756 0.505nn 0.547nn 1 Oper. 3.699 0.560 0.367nn 0.270nn 0.311nn 1 perf. nn 0.071 1 Age 38.974 27.712 0.490 0.100 0.210 Size 817.120 2840.044 0.028 0.156n 0.065 0.093 0.004
n

nn

p o0.05. po 0.01.

A. Furlan et al. / Int. J. Production Economics 133 (2011) 489495

493

Table 4 OLS regression: IntJIT and UpJIT on operational performance. Variables High IntJIT&UpJIT Mainly IntJIT Mainly UpJIT Low IntJIT&UpJIT Age Size Coefcient 3.848*** 3.845*** 3.598*** 3.589*** 0.001 0.000 Std. error 0.085 0.106 0.114 0.086 0.001 0.000 t-value 45.072 36.201 31.535 41.901 0.749 1.516

5. Discussion Literature on world class manufacturing highlights the importance of investing on different initiatives to improve performance. In particular, OM literature has proved that the implementation of internal and external JIT practices improves the overall performance of the rm and its competitive advantage (Flynn et al., 1999; Frohlich and Westbrook, 2001; Shah and Ward, 2007) However, we maintain that the synergistic effects among internal and external JIT practices deserve further empirical validation. In this paper, we rely on the theory of complementarity (Milgrom and Roberts, 1990, 1995) to shed light on the complex and tangled relations among internal and external JIT bundles across the supply chain. Grounded on the literature, we develop a set of testable hypotheses on the complementary effects between internal JIT, upstream JIT (towards suppliers) and downstream JIT (towards customers). We then use statistical analyses on High Performance Manufacturing round III database to test these hypotheses. Our results suggest that, to maximize operational performance, different approaches should be adopted in implementing internal JIT, upstream JIT and downstream JIT. Firstly, we nd that complementarity does not hold either between internal JIT and upstream JIT or between internal JIT and downstream JIT. From a theoretical perspective, the lack of complementarity points towards a modular-wise design approach to internal and external JIT bundles. Since internal and external JIT bundles do not produce enough synergistic effects to generate complementarity, they can be designed and implemented as modules. In other words, the interdependencies between internal and external JIT bundles are not so complex to inhibit the organizational advantages of decoupling (Simon, 1962; Sanchez and Mahoney, 1996; Schilling and Steensma, 2001). These advantages are related to the fact that the whole improvement process can be decomposed into simpler and smaller tasks that are easier to be tackled and implemented by the rm. A modular approach in designing JIT bundles should be aimed at reducing the interdependencies between internal and external bundles while creating standard interfaces (Ulrich, 1995; Salvador et al., 2002) to connect internal and external upstream and downstream logistics ows. Examples of standard interfaces involving internal and external JIT bundles could be the adoption of standard pallets, trays or containers, the development of formal reports to be lled by the material handlers, the denition of standard procedures to be followed to implement vendor Kanban, etc. From a managerial perspective, these results are coherent with the literature suggesting a sequential approach in implementing integration practices (Stevens, 1989; Cooper and Ellram, 1993; Cagliano et al., 2006). These authors argue that the implementation process should start with internal integration and then involve both suppliers and customers (Scott and Westbrook, 1991; Cooper and Ellram, 1993). Indeed, since internal and upstream/downstream JIT are not complementary, it is likely that a simultaneous implementation of these bundles do not produce enough benets to offset the associated costs. Conversely, to facilitate the implementation process, a rm should be better off if it decides to implement the different modules sequentially adopting an incremental approach. On the whole, these results support the widespread practice of implementing rst internal JIT bundles and then external JIT bundles. Firms have long learned that time, resource and competency constraints limit their ability to manage too highly complex improvement initiatives, e.g. those that simultaneously involve internal and external JIT bundles.

Complementarity test between IntJIT and UpJIT. High IntJIT&UpJIT Mainly IntJIT4Mainly UpJIT Low IntJIT&UpJIT. F(1.203) 0.003, N 266, p-value 0.954. Coefcients signicant at 1%***, 5%** and 10%*.

Table 5 OLS regression: IntJIT and DownJIT on operational performance. Variables High IntJIT&DownJIT Mainly IntJIT Mainly DownJIT Low IntJIT&DownJIT Age Size Coefcient 3.861*** 3.796*** 3.739*** 3.502*** 0.001 0.000 Std. error 0.085 0.104 0.106 0.088 0.001 0.000 t-value 45.370 36.590 35.421 40.016 0.476 1.240

Complementarity test between IntJIT and DownJIT. High IntJIT&DownJIT Mainly IntJIT 4Mainly DownJIT Low IntJIT&DownJIT. F(1.203) 1.180, N 266, p-value 0.279. Coefcients signicant at 1%***, 5%** and 10%*.

Table 6 OLS regression: UpJIT and DownJIT on operational performance. Variables High UpJIT&DownJIT Mainly UpJIT Mainly DownJIT Low UpJIT&DownJIT Age Size Coefcient 3.889*** 3.529*** 3.704*** 3.640*** 0.001 0.000 Std. error 0.087 0.111 0.106 0.088 0.001 0.000 t-value 44.685 31.893 35.076 41.581 0.366 0.878

Complementarity test between UpJIT and DownJIT. High UpJIT&DownJIT Mainly UpJIT4Mainly DownJIT Low UpJIT&DownJIT. F(1.203) 3.237*, N 266, p-value 0.074. Coefcients signicant at 1%***, 5%** and 10%*.

Low UpJIT&DownJIT are highly signicant. As for the control variables, rm age and size do not have a signicant impact on operational performance. We use the following test for complementarity: y11 y10 Z y01 y00 (where y11 is the coefcient/marginal return associated to the High UpJIT&DownJIT group, y10 is the coefcient/marginal return associated to the Mainly UpJIT group, y01 is the coefcient/ marginal return associated to the Mainly DownJIT group and y00 is the coefcient/marginal return associated to Low UpJIT&DownJIT). Using F-test, we accept the complementarity (p-value0.074). In line with the fact that upstream JIT and downstream JIT are complements, we nd a positive correlation between these two bundles (r 0.547, po0.01) (see Table 3). Further evidence supporting complementarity between downstream JIT and upstream JIT can be found using the frequency with which rms combine these JIT bundles. The two groups of rms reporting the highest adoption frequency are High UpJIT&DownJIT and Low UpJIT&DownJIT (Table 7). This result signals that rms tend to adopt a high level of upstream JIT and downstream JIT or a low level of both. This is coherent with the general assertion that two complementary activities are positively correlated.

494

A. Furlan et al. / Int. J. Production Economics 133 (2011) 489495

Table 7 Frequency and mean of lean bundles by operational performance. Lean strategy High UpJIT&DownJIT Mainly UpJIT Mainly DownJIT Low UpJIT&DownJIT Frequency 76 38 43 88 (31%) (15%) (18%) (36%) Mean of performance 3.9243 3.4868 3.7054 3.5919 Tukey differences Mainly UpJIT**, Low UpJIT&DownJIT** High UpJIT&DownJIT** High UpJIT&DownJIT**

The differences in means are signicant: F(3.241) 7.57, p-value 0.000. Coefcients signicant at 5%**. This sample (N 245) only includes rms that reported non-missing observations on all variables used in the analysis.

Our second result is that upstream JIT and downstream JIT bundles are complementary. The complementarity between upstream and downstream JIT provides robust conrmation to the arc of integration theory (Frohlich and Westbrook, 2001), suggesting that the widest integration arc involving both suppliers and customers leads to the highest rm performance (Vereecke and Muylle, 2006). Coherent design of upstream and downstream JIT lean bundles reduces uncertainty and variability (Lee et al., 1997), and minimizes waste along the supply chain (Hall, 1987). The existence of complementarity between upstream and downstream JIT has relevant managerial implications too. We argue that the presence of complementarity suggests that a concurrent approach in designing (and possibly implementing) upstream and downstream JIT should be adopted if the rm is seeking to maximize operational performance. The fact that upstream and downstream JIT are complementary implies that these two bundles share complex interdependencies (Milgrom and Roberts, 1995). This suggests that logistic initiatives aimed at streamlining the whole supply chain require a tight coordination between upstream and downstream JIT to capture the positive effects that each JIT bundle has on the marginal returns of the other. We maintain that this coordination is achieved by means of both process standardization and mutual adaptation (Thompson, 1967). The rst mechanism requires a tight coordination in designing these two bundles to allow the introduction of a common set of routines and procedures that each bundle of activities has to follow. The second mechanism requires informal interactions between the two bundles to achieve mutual adjustments over time. We believe that a simultaneous implementation of both upstream and downstream JIT bundles is likely to facilitate the formation of those activities links (Hakansson et al., 2009) that are necessary to govern mutual adjustments over time. We nally argue that improvement initiatives simultaneously involving both upstream and downstream JIT bundles are not too complex for the rms to manage. Indeed, we expect that the rms that are ready to implement upstream and downstream JIT bundles have already implemented internal JIT bundle. These rms therefore have accumulated valuable organizational competencies in managing lean projects moving down the learning curve and lowering the learning barriers to implementing improvement programs. These rms are likely to have accumulated a stock of capabilities which allow them to manage improvement projects characterized by a high level of complexity.

while we use an overall performance measure that derives from the average of four individual performance dimensions, future studies could test the existence of complementarity considering single performance areas (cost, quality, exibility and time). Other methodologies can be adopted to rene the complementarity analysis. For example, MacDufe (1995) considers the interaction terms in a regression analysis to test the complementarity between different bundles of practices. Along the same vein, recent economic studies on complementarity (Miravete and Pernias, 2006) have introduced methodological approaches able to infer the correlation introduced by unobserved heterogeneity and distinguish it from the complementarity. Future studies could also extend the test of complementarity to other internal and external lean bundles (e.g. total quality management and human resource management). Finally, future research can detect whether the position of the focus rm within the supply chain affects the complementarity between internal and external lean bundles.

References
Athey, S., Stern, S., 1998. An empirical framework for testing theories about complementarity in organizational design. National Bureau of Economic Research Working Paper 6600. Ballou, R.-H., 1992. Business Logistics Management 3rd Edition Prentice-Hall, Englewood Cliffs, NJ. Bocquet, R., Brossard, O., Sabatier, M., 2007. Complementarities in organizational design and the diffusion of information technologies: an empirical analysis. Research Policy 36, 367386. Bowersox, D.-J., 1989. Logistics in the integrated enterprise. The Annual Conference of the Council of Logistics Management, St. Louis, MO. Cagliano, R., Caniato, F., Spina, G., 2006. The linkage between supply chain integration and manufacturing improvement programmes. International Journal of Operations & Production Management 26 (3), 282299. Cassiman, B., Veugelers, R., 2006. In search of complementarity in innovation strategy. Management Science 52 (1), 6882. Cooper, M.-C., Ellram, L.-M., 1993. Characteristics of supply chain management and implications for purchasing and logistics strategy. The International Journal of Logistics Management 4 (2), 1322. Cooper, M.-C., Lambert, D.-M., Pagh, J.-D., 1997. Supply chain management: more than a new name for logistics. The International Journal of Logistics Management 8 (1), 113. Dal Pont, G., Furlan, A., Vinelli, A., 2008. Interrelationships among lean bundles and their effects on operational performance. Operations Management Research 1 (2), 150158. Das, A., Narasimhan, R., Talluri, S., 2006. Supplier integrationnding an optimal conguration. Journal of Operations Management 24, 563582. Dong, Y., Carter, C.-R., Dresner, M.-E., 2001. JIT purchasing and performance: an exploratory analysis on buyer and supplier perspectives. Journal of Operations Management 19, 471483. Ettlie, J.-E., Stoll, H.-W., 1990. Managing the DesignManufacturing Process. McGraw-Hill, New York. Flynn, B.-B., Schroeder, R.-G., Flynn, G.-E., 1999. World class manufacturing: an investigation of Hayes and Wheelwrights foundation. Journal of Operations Management 17, 249269. Frohlich, M.-T., Westbrook, R., 2001. Arcs of integration: an international study of supply chain strategies. Journal of Operations Management 19 (2), 185200. Gimenez, C., Ventura, E., 2005. Logistics-production, logistics-marketing and external integration: their impact on performance. International Journal of Operations & Production Management 25 (1), 2038. Hahn, C.-K., Pinto, P.-A., Bragg, D.-J., 1983. Just-in-time production and purchasing. International Journal of Purchasing and Materials Management 19 (4), 210.

6. Limitations and future research Our study suffers from some limitations that can provide indications for future research. Even if we provide fresh knowledge about the complementarity between JIT bundles adopted along the supply chain, more studies are needed to achieve a full understanding of such a complementarity. A longitudinal approach can be adopted to better examine the causal relationships among practices and performance. Moreover,

A. Furlan et al. / Int. J. Production Economics 133 (2011) 489495

495

Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate Analysis. Prentice Hall International, Englewood. Hakansson, H., Ford, D., Gadde, L., Snehota, I., Waluszewski, A., 2009. Business in Networks. Wiley, Chichester. Hall, R.W., 1987. Attaining Manufacturing Excellence: Just-in-time, Total Quality, Total People InvolvementDow Jones-Irwin, Homewood, IL. Harland, C.-M., Lamming, R.-C., Cousins, P.-D., 1999. Developing the concept of supply strategy. International Journal of Operations & Production Management 19 (7), 650674. Helper, S., 1991. How much has really changed between US automakers and their suppliers? Sloan Management Review 15 1528. Hill, T., 1994. Manufacturing Strategy: Text and Cases. Homewood, IL, Irwin. Holweg, M., 2007. The genealogy of lean production. Journal of Operations Management 25, 420437. Kulp, S.-C., Lee, H.-L., Ofek, E., 2004. Manufacturers benets from information integration with retail customers. Management Science 50 (4), 431444. Lamming, R.-C., 1993. Beyond Partnership: Strategies for Innovation and Lean Supply. Prentice-Hall, Englewood Cliffs, NJ. Lee, H.-L., Billington, C., 1992. Managing supply chain inventory: pitfalls and opportunities. Sloan Management Review, 6573. Lee, H.-L., Padmanabhan, V., Whang, S., 1997. Information distortion in a supply chain: the bullwhip effect. Management Science 43 (4), 546558. MacDufe, J.P., 1995. Human resource bundles and manufacturing performance: organizational logic and exible production systems in the world auto industry. Industrial and Labor Relations Review 48 (2), 197221. McLachlin, R., 1997. Management initiatives and just-in-time manufacturing. Journal of Operations Management 15, 271292. Mefford, R., 1989. The productivity nexus of new inventory and quality control techniques. Engineering Costs and Production Economics 17, 2128. Milgrom, P., Roberts, J., 1990. The economics of modern manufacturing: technology, strategy and organization. American Economic Review 80, 511528. Milgrom, P., Roberts, J., 1995. Complementarities and t: strategy, structure and organizational change in manufacturing. Journal of Accounting and Economics 19, 179208. Miravete, E., Pernias, J., 2006. Innovation complementarity and scale of production. Journal of Industrial Economics 54, 129. Narasimhan, R., Das, A., 2001. The impact of purchasing integration and practices on manufacturing performance. Journal of Operations Management 19 (5), 593609. Olsen, R.-F., Ellram, L.-M., 1997. A portfolio approach to supplier relationships. Industrial Marketing Management 26, 101113. Pagell, M., Handeld, R.-B., Barber, A.-E., 2000. Effects of operational employee skills on advanced manufacturing technology performance. Production and Operations Management 9 (3), 222238. Rosenzweig, E.-D., Roth, A.-V., Dean, J.-W., 2003. The inuence of an integration strategy on competitive capabilities and business performance: an exploratory study of consumer product manufacturers. Journal of Operations Management 21, 437456. Sazadeh, M.-H., Ritzman, L.-P., Sharma, D., Wood, D., 1996. An empirical analysis of the product process matrix. Management Science 42 (11), 15761591. Salvador, F., Forza, C., Rungtusanatham, M., 2002. Modularity, product variety, production volume, and component sourcing: theorizing beyond generic prescriptions. Journal of Operations Management 20, 549575.

Sanchez, R., Mahoney, J.T., 1996. Modularity, exibility, and knowledge management in product and organization design. Strategic Management Journal 17, 6376. Schilling, M., Steensma, H., 2001. The use of modular organizational forms: an industry-level analysis. Academy of Management Journal 44 (6), 11491168. Schroeder, R.-G., Flynn, B.-B., 2001. High Performance Manufacturing: Global Perspectives. Wiley, New York, NY. Scott, C., Westbrook, R., 1991. New strategic tools for supply chain management. International Journal of Physical Distribution & Logistics Management 21 (1), 2333. Shah, R., Ward, P.-T., 2003. Lean manufacturing: context, practice bundles and performance. Journal of Operations Management 21, 129149. Shah, R., Ward, P.-T., 2007. Dening and developing measures of lean production. Journal of Operations Management 25, 785805. Siggelkow, N., 2001. Change in the presence of t: the rise, the fall, and the renaissance of Liz Clairborne. Academy of Management Journal 44, 838857. Siggelkow, N., 2002a. Evolution toward t. Administrative Science Quarterly 47, 125159. Siggelkow, N., 2002b. Misperceiving interactions among complements and substitutes: organizational consequences. Management Science 48, 900916. Simchi-Levi, D., Kaminsky, P., Simchi-Levi, E., 2003. Designing and Managing the Supply Chain. Concepts, Strategies and Case Studies. McGraw-Hill/Irwin, Boston, MA. Simon H.A., 1962. The architecture of complexity. In: Proceedings of the American Philosophical Society, vol. 106, pp. 467482. Slack, N., Chambers, S., Johnston, R., 2007. Operations Management. Pitman, London. Sriparavatsu, L., Gupta, T., 1997. An empirical study of just-in-time and total quality management principles implementation in manufacturing rms in the USA. International Journal of Operations & Production Management 17 (12), 12151232. Stank, T.-P., Keller, S., Daugherty, P.-J., 2001. Supply chain collaboration and logistical service performance. Journal of Business Logistics 22 (1), 2948. Stevens, G.-C., 1989. Integrating the supply chain. International Journal of Physical Distribution & Materials Management 19 (8), 38. Tan, K.-C., Kannan, V.-R., Handeld, R.-B., 1998. Supply chain management: supplier performance and rm performance. International Journal of Purchasing and Materials Management 34 (3), 29. Thompson, J.D., 1967. Organization in Action. McGraw-Hill, New York. Ulrich, K.-T., 1995. The role of product architecture in the manufacturing rm. Research Policy 24, 419440. Vereecke, A., Muylle, S., 2006. Performance improvement through supply chain collaboration in Europe. International Journal of Operations & Production Management 26 (11), 11761198. Verma, R., Thompson, G.-M., Moore, W.-L., Louvriere, J.-J., 2001. Effective design of product/services: an approach based on integration of marketing and operations management decision. Decision Sciences 31 (1), 165193. Watts, C.-A., Kim, Y.-K., Hahn, C., 1992. Linking purchasing to corporate competitive strategy. International Journal of Purchasing and Materials Management 28 (4), 28. Womack, J.-P., Jones, D.-T., 1996. Lean Thinking: Banish Waste and Create Wealth in Your Corporation. Simon & Schuster, New York. Womack, J.-P., Jones, D.-T., Roos, D., 1990. The Machine that Changed the World. Rawson Associates, New York, NY.