Journal of Operations Management 29 (2011) 343355

Contents lists available at ScienceDirect

Journal of Operations Management

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

Agile manufacturing: Relation to JIT, operational performance and rm performance

R. Anthony Inman a, , R. Samuel Sale b , Kenneth W. Green Jr. c,1 , Dwayne Whitten d,2
a

College of Business, Louisiana Tech University, Box 10318, Ruston, LA 71272, United States Department of Management and Marketing, PO Box 10025, Lamar University, Beaumont, TX 77710, United States c Department of Management, Marketing, and MIS, College of Business, Southern Arkansas University, P.O. Box 9410, Magnolia, AR 71754, United States d Texas A&M University - Mays Business School, Information and Operations Management Department, Mailstop 4217, College Station, TX 77843, United States
b

a r t i c l e

i n f o

a b s t r a c t
A structural model incorporating agile manufacturing as the focal construct is theorized and tested. The model includes the primary components of JIT (JIT-purchasing and JIT-production) as antecedents and operational performance and rm performance as consequences to agile manufacturing. Using data collected from production and operations managers working for large U.S. manufacturers, the model is assessed following a structural equation modeling methodology. The results indicate that JIT-purchasing has a direct positive relationship with agile manufacturing while the positive relationship between JITproduction and agile manufacturing is mediated by JIT-purchasing. The results also indicate that agile manufacturing has a direct positive relationship with the operational performance of the rm, that the operational performance of the rm has a direct positive relationship with the marketing performance of the rm, and that the positive relationship between the operational performance of the rm and the nancial performance of the rm is mediated by the marketing performance of the rm. 2010 Elsevier B.V. All rights reserved.

Article history: Received 7 January 2007 Received in revised form 1 June 2010 Accepted 5 June 2010 Available online 18 June 2010 Keywords: Agile manufacturing JIT systems Organizational performance Structural equation modeling

1. Introduction Competitive pressures force manufacturers to continuously improve the provision of products and associated services desired by customers. Manufacturers have adopted lean practices such as JIT and TQM to reduce costs and improve quality. As many competitors adopted these practices, some competitive advantage was lost. Many manufacturers now have begun adopting practices that increase their ability to rapidly respond to changes in customer demand. For these, superior responsiveness has become a key to competitive advantage. In short, many manufacturing rms are becoming relatively more agile. We propose that an element of lean manufacturing, Just-in-Time (JIT), is related to agile manufacturing. Specically, we propose that the primary elements of JIT, i.e., JIT-production and JIT-purchasing, are related to agility. Further we investigate the relationship between manufacturing agility and operational and rm performance.

We conducted a national survey of production and operations managers working for large U.S. manufacturing concerns to collect data necessary to assess the model using a structural equation methodology. A review of the literature and discussion of the study hypotheses follow in the next section. A discussion of the specic methodology employed is followed by a description of the results of the scale assessment and the structural equation modeling results. Finally, a conclusions section, which incorporates discussions of the contributions of the study, limitations of the study, suggestions for future related research, and implications for practicing managers, is provided. 2. Literature review and hypotheses Shah and Ward (2003) identify JIT as one of four bundles that make up lean manufacturing. Given that JIT is an element of lean manufacturing, discussion of the literature relating lean manufacturing to agile manufacturing is relevant even though the current study focuses on the relationship between JIT and agile manufacturing. Hence, the following section provides a review of the literature for both the JIT/agile relationship and the lean/agile relationship. 2.1. JIT and agile manufacturing Specic to our research is the relationship between agile manufacturing and the Just-in-Time (JIT) manufacturing strategy.

Corresponding author. Tel.: +1 318 257 3568; fax: +1 318 257 4253. E-mail addresses: inman@latech.edu (R.A. Inman), sam.sale@lamar.edu (R.S. Sale), kwgreen@saumag.edu (K.W. Green Jr.), dwhitten@mays.tamu.edu (D. Whitten). 1 Tel.: +1 870 235 4317 (O). 2 Tel.: +1 979 845 2919 (O). 0272-6963/$ see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jom.2010.06.001

344

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

Countless research regarding JIT and its individual elements has been generated in the last three decades. Claycomb et al. (1999b) state that in its ideal form, JIT integrates the entire supply chains marketing, distribution, customer service, purchasing, and production functions into one controlled process. In an early work regarding JIT implementation, Mehra and Inman (1992) identied four elements of JIT: JIT-production strategy, JIT vendor strategy (purchasing), JIT education strategy and management commitment. Only JIT-production and JIT vendor strategies were found to have a signicant impact on JIT implementation success. Since that time a number of published articles have at least partially supported these ndings. In more recent work Shah and Ward (2003) identify four bundles of lean production: Just-In-time (JIT), Total Quality Management (TQM), Total Preventive Maintenance (TPM) and Human Resource Management (HRM). In a 2007 paper Shah and Ward propose and test 10 dimensions that can be used to measure these four bundles of lean production. Six of the 10 dimensions are elements of JIT with three pertaining to supplier aspects of JIT (purchasing) and three related to aspects of JITproduction. Therefore, while a number of JIT elements have been identied, two, JIT-production and JIT-purchasing, seem to garner the most support for their criticality to organization success. As a result, we limit our work here to these two primary elements of JIT. We dene JIT as a comprehensive strategy that combines the primary tactical elements of JIT-production and JIT-purchasing, to eliminate waste and optimally utilize resources throughout the supply chain (Claycomb et al., 1999b). JIT-production focuses on the identication and elimination of all forms of waste, including excess inventories, material movements, production steps, scrap losses, rejects and rework, within the production function (Wisner et al., 2005; Brox and Fader, 2002). JIT-purchasing is operationalized by Freeland (1991) as a set of techniques and concepts for eliminating waste and inefciency in the purchasing process. Techniques and concepts associated with JIT-purchasing include daily delivery of small lot sizes from nearby vendors, shared information, supplier education, reduced inspection and early supplier involvement in product/process design. The techniques utilized by JIT-production and JIT-purchasing allow rms to translate the resulting capabilities into a JIT strategy that provides organizational capabilities to deliver near zero defect quality, near zero variance quantity and precise on-time delivery (Green and Inman, 2005). The key word applicable to the denition of both primary elements of JIT is waste. This is consistent with Shah and Wards (2007) denition of lean production as an integrated sociotechnical system with the main objective of reducing or eliminating internal, customer, and supplier waste. Since JIT is a subset (bundle) of lean, we narrow our denition to the following: JIT is that subset of lean associated primarily with the elimination of waste through planning, scheduling and sequencing of operations. This denition of JIT subsumes both primary elements of JIT, JIT-purchasing and JIT-production, as elements of itself that are distinguishable from each other by where they occur in the system or supply chain. 2.2. Lean manufacturing and agile manufacturing There has been a tendency to view the development of lean manufacturing and agile manufacturing either in a progression or in isolation (Gunasekaran, 1999a). From an isolation standpoint, Harrison (1997) notes that companies with a lean mindset would nd the agile manufacturing concept difcult to follow. Krishnamurthy and Yauch (2007) state that there are three general positions with respect to lean and agile: those who believe that they are mutually exclusive or distinct concepts that cannot co-exist, those who believe that they are mutually supportive strategies, and those who believe that leanness must be a precursor to agility. Table 1 summarizes the literature supporting each of the three views.

2.2.1. Lean and agile as mutually exclusive concepts Early concerns that the two concepts cannot co-exist were expressed by Richards (1996), who noted that some agile proponents claimed that exibility would suffer under lean production and from Harrison (1997) who expressed doubts that lean and agile were compatible while emphasizing that agile implied more resources, not fewer. More recently, Goldsby et al. (2006) note that lean and agile are often pitted as opposing paradigms. Agility has been recognized as a manufacturing strategy consisting of manufacturing tasks and choices (Gunasekaran et al., 2008). The word choices implies that tradeoffs are necessary between lean and agile (Harrison, 1997) or that they cannot completely coexist. While both strategies address the same competitive priorities (cost, quality, service, exibility), they each emphasize different elements (Narasimhan et al., 2006) such that clear dividing lines can be drawn between the two (Gunasekaran et al., 2008). Some would state that lean manufacturing subordinates responsiveness (service) to efciency and productivity (cost) (Vazquez-Bustelo et al., 2007) while agile manufacturing focuses on speed and exibility and not cost (Gunasekaran et al., 2008). One may consider leans market winner as cost (Christopher and Towill, 2001) and agiles market winners as speed, exibility and responsiveness to changes (Zhang and Shari, 2007), i.e., service level (Mason-Jones et al., 2000). This is consistent with Narasimhans et al. (2006) empirical study that found agile plants to meet/exceed lean plants and other plants in all measured performance dimensions with the exception of cost efciency. Hence, tradeoffs that would prevent lean/agile co-existence can be easily envisioned. Larger lot sizes and higher inventory levels could be necessary to maintain the higher service level required by agile rms while smaller lot sizes and lower inventory levels could be required by cost-efcient lean rms. It should be noted that there is a stream of thought that advocates the simultaneous use of lean manufacturing and agile manufacturing. Termed leagile, proponents believe that manufacturing systems can consist of both lean and agile, acting together to exploit market opportunities in a cost-efcient manner (Krishnamurthy and Yauch, 2007). However, this appears to be appropriate only for supply chains, not individual manufacturing rms unless the rm is a multi-unit enterprise that functions as a supply chain. Leagile models created thus far contain a decoupling point that separates the lean and agile portions of the system (Krishnamurthy and Yauch, 2007) with the lean portion on the upstream side of the point and the agile portion of the system on the downstream side (Mason-Jones et al., 2000). Krishnamurthy and Yauch (2007) state that this decoupling point ensures that lean and agile do not co-exist, lending credence to the idea that the two are mutually exclusive within a single manufacturing entity, although both may exist within a supply chain. From the literature, one can glean that both lean and agile have obtained desired results in isolation and that neither is better nor worse than the other (Naylor et al., 1999). This would imply that either could be used successfully depending upon the individual rms environment. Specically, lean manufacturing is appropriate when market conditions are basically stable, demand is smooth and standard products are produced and agile manufacturing is appropriate when the environment is more turbulent and more product variety is present (Vazquez-Bustelo et al., 2007; Naylor et al., 1999). The degree of turbulence in the environment determines the degree of agility needed (Vazquez-Bustelo et al., 2007; Shari and Zhang, 2001; Zhang and Shari, 2000). Though not stated within the literature, the same could hold true for lean. The degree of stability dictates the degree of leanness required to effectively compete. Consistent with the above, Goldsby et al. (2006) found, via simulation, that a lean strategy resulted in the lowest cost/highest service when demand was smooth and predicted with a high degree of accuracy coupled with low-value nished goods

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355 Table 1 Three views of the relationship between lean (JIT) and agile manufacturing. Lean and agile as mutually exclusive concepts Harrison (1997) Goldsby et al. (2006) Narasimhan et al. (2006) Gunasekaran et al. (2008) Vazquez-Bustelo et al. (2007) Christopher and Towill (2001) Zhang and Shari (2007); Mason-Jones et al. (2000) Vazquez-Bustelo et al. (2007); Naylor et al. (1999)

345

Expressed doubts that lean and agile were compatible while emphasizing that agile implied more resources, not fewer Note that lean and agile are often pitted as opposing paradigms They each emphasize different elements Clear dividing lines can be drawn between the two; agile manufacturing focuses on speed and exibility and not cost Lean manufacturing subordinates responsiveness (service) to efciency and productivity (cost) Leans market winner is cost Agiles market winners are speed, exibility and responsiveness to changes, i.e., service level Lean manufacturing is appropriate when market conditions are basically stable, demand is smooth and standard products are produced and agile manufacturing is appropriate when the environment is more turbulent and more product variety is present Leanness is an overarching concept that is compatible with any production system Mutually supportive concepts Results in benets not accessible when the concepts are used in isolation Compatible concepts Complementary concepts Elements cited as necessary for agile performance include elements of lean manufacturing, specically Just-in-Time manufacturing The predominant view in the literature is that lean manufacturing is a performance/practice state that is antecedent to agile manufacturing Agile is the latest step in the evolution from mass production, to Just-in-Time to lean to agile Agile manufacturing assimilates the full range of exible production technologies, along with the lessons learned from TQM, JIT, and lean production Agile manufacturing can be achieved by utilizing and integrating elements of existing systems and methods that are already developed and in use Agile manufacturing = exible manufacturing system + lean manufacturing Agile manufacturing can subsume the paradigm of lean production Agile manufacturing is the next logical step or a natural development from the concept of lean manufacturing

Lean and agile as mutually supportive concepts Katayama and Bennett (1999) Katayama and Bennett (1999); Krishnamurthy and Yauch (2007) Krishnamurthy and Yauch (2007) Kidd (1994) Naylor et al. (1999) Gunasekaran et al. (2008); Ramesh and Devadasan (2007); Goldsby et al. (2006): McCullen and Towill (2001) Lean as antecedent to agility Narasimhan et al. (2006) Jin-Hai et al. (2003); Hormozi (2001) Goldman and Nagel (1993) Gunasekaran et al. (2008); Vazquez-Bustelo et al. (2007); Shari and Zhang (2001); Zhang and Shari (2000) Sarkis (2001) McCullen and Towill (2001) Gunasekaran et al. (2008); Hormozi (2001); Maskell (2001); Gunasekaran (1999b); Robertson and Jones (1999); Booth (1996)

and low carrying costs. Results of a study by Narasimhan et al. (2006) indicated that lean performers had made-to-stock operations while agile performers had a signicantly greater proportion of to-order operations. 2.2.2. Lean and agile as mutually supportive concepts Alternately, leanness has been described as an overarching concept that is compatible with any production system (Katayama and Bennett, 1999) and as such should be compatible (Krishnamurthy and Yauch, 2007; Kidd, 1994), complementary (Naylor et al., 1999), and mutually supportive (Krishnamurthy and Yauch, 2007; Katayama and Bennett, 1999) with agile manufacturing, resulting in benets not accessible when the concepts are used in isolation (Krishnamurthy and Yauch, 2007). Elements cited as necessary for agile performance include: the ability to produce large or small batches with minimum setups (and setup time) and a cross-trained exible workforce (Goldsby et al., 2006); reduced process lead times and costs (Gunasekaran et al., 2008); relationships with suppliers and JIT-production (McCullen and Towill, 2001); fully empowered employees, JIT-purchasing, and exible setups (Ramesh and Devadasan, 2007). Interestingly, these are all elements of lean manufacturing, specically Just-in-Time manufacturing. Based on the above logic, it seems that the two concepts could indeed be mutually supportive. 2.2.3. Lean as antecedent to agility A number of researchers feel that agile manufacturing can be achieved by utilizing and integrating elements of existing systems and methods that are already developed and in use (Gunasekaran et al., 2008; Vazquez-Bustelo et al., 2007; Shari and Zhang, 2001; Zhang and Shari, 2000). More specically, there are those that feel that agile manufacturing is the next logical step or a natural devel-

opment from the concept of lean manufacturing (Gunasekaran et al., 2008; Hormozi, 2001; Maskell, 2001; Gunasekaran, 1999b; Robertson and Jones, 1999; Booth, 1996). Sarkis (2001) offers the formula: agile manufacturing = exible manufacturing system + lean manufacturing. McCullen and Towill (2001) argue that agile manufacturing can subsume the paradigm of lean production. Specic, to our research, Narasimhan et al. (2006) report that the predominant view in the literature is that lean manufacturing is a performance/practice state that is antecedent to agile manufacturing, with results of their study suggesting that leanness is a precursor to agility. One may summarize this part of the literature review with Goldman and Nagels (1993) statement that agile manufacturing assimilates the full range of exible production technologies, along with the lessons learned from TotalQuality-Management [an element of lean, Shah and Ward, 2003], Just-in-Time production [an element of lean, Shah and Ward, 2003] and lean production. Perusing the literature review also begs the question, when is lean manufacturing assimilated into the agile system? Does it have to be established before moving on to agile manufacturing, can a lean system be established at the same time and as a part of an agile system, or does it really matter? A number of researchers state that agile is the latest step in the evolution from mass production, to Just-in-Time to lean to agile (Jin-Hai et al., 2003; Hormozi, 2001). If this is the case, then most agile rms probably adopted lean at some point and then later moved on to agile, making lean a precursor to agile. Simply stated by Narasimhan et al. (2006), results indicate that while the pursuit of agility might presume leanness, pursuit of leanness might not presume agility. Since most of the evidence put forth by the precursor literature would just as well justify a mutually supportive stance, we make the assumption that if lean is antecedent to agile, as proposed,

346

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

mutual support will also be present. This leaves us with two options (1) lean manufacturing (JIT) is antecedent to agile manufacturing (with mutual support assumed), i.e., higher levels of JIT will result in higher levels of agility, or (2) they are distinct concepts that cannot co-exist, that is, the increased effectiveness in one area results in a decrease in effectiveness in the other. 2.3. Hypotheses As stated earlier, Shah and Ward (2003) identied JIT as one of the four bundles that make up lean manufacturing so the preceding discussion involving lean manufacturing is assumed to apply also to the specic lean bundle, JIT. This assumption is supported by Vazquez-Bustelo et al. (2007) who state that experience suggests a JIT-production system is required for agility and Narasimhan et al. (2006) who found that supplier management and JIT ow and layout received a signicantly higher emphasis in agile rms than in lean rms. Also, as previously noted, many elements cited as necessary for agile performance are previously established elements of JIT. Hence, our research question becomes Is JIT antecedent to agile manufacturing (thus, the two are mutually supportive) or are JIT and agile manufacturing two distinct concepts that cannot coexist? Using the two primary elements of JIT, we propose two hypotheses to dene our research objective. H1. Higher levels of adoption of a JIT-purchasing strategy will lead to higher levels of a rms manufacturing agility, i.e., JIT-purchasing is antecedent to agility. H2. Higher levels of adoption of a JIT-production strategy will lead to higher levels of a rms manufacturing agility, i.e., JIT-production is antecedent to agility. Hypotheses 1 and 2 are conceptually pictured in Fig. 1. If test results indicate a signicant negative relationship between the JIT strategies and agile manufacturing, a mutually exclusive relationship between the two will be supported. Manufacturers become more agile with the expectation of improving performance (Yusuf and Adeleye, 2002; Mason-Jones et al., 2000). Organizational performance encompasses both nancial and marketing performance at the rm level (Green and Inman, 2005; Green et al., 2004). Financial performance focuses on a rms

return on investment, return on sales and protability as compared to its competition. The marketing performance component compares the rms sales volume, sales growth, and market share to that of its competition. Yusuf and Adeleye (2002) surveyed 109 manufacturers and found a signicant link between agility and business performance (sales turnover, market share, customer loyalty, performance relative to competitors, and aggregate performance). Vazquez-Bustelo et al. (2007) surveyed rms in Spain and found that agile manufacturing positively impacted manufacturing strength which led to improved operational, nancial and market performance. Results of a survey by Narasimhan et al. (2006) revealed that agile plants met or exceeded lean and other plants in all measured performance dimensions except cost efciency, giving agility the appearance of a higher state of plant performance and capability. We propose that rms adopting agile manufacturing practices will experience improved operational and rm performance. The following hypotheses were fashioned based upon this proposition: H3. Higher levels of manufacturing agility will have a positive impact on a rms nancial performance. H4. Higher levels of manufacturing agility will have a positive impact on a rms marketing performance. H5. Higher levels of manufacturing agility will have a positive impact on a rms operational performance. Hypotheses 35 are also conceptually pictured in Fig. 1. 3. Methodology A listing of 1350 plant and operations managers was extracted from the 2004 Manufacturers News, incorporated database of U.S. manufacturers with more than 250 employees. Plant and operations managers were targeted because of their particular knowledge related to the manufacturing processes within their organizations. It was assumed that plant and operations managers, as a group, would be interested in participating in the survey and would readily understand the survey items. The survey instrument was moderately long, lling the front and back of two legal-size pages.

Fig. 1. Agile manufacturing model with hypotheses.

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

347

Each of the manufacturers was mailed an initial request to participate that included a cover letter, a non-participating form, the survey instrument, and a postage-paid return envelope. The cover letter requested participation and stated an assurance that all responses would be anonymous. The non-participating form allowed plant managers who did not wish to participate in the study to have their names and addresses removed from the database. A follow-up mailing that included a revised cover letter, another survey instrument, and return envelope was completed 2 weeks after the initial mailing. This second mailing did not include managers who lled out the non-participating form. A descriptive prole of respondents was prepared, and early and late responders were compared to assess for non-response bias. Conrmatory factor analysis was used to assess the dimensionality of the study scales, and all scales were further assessed for reliability and validity. Summary values were computed for each study variable. Descriptive statistics for each of the variables were computed and a correlation matrix prepared. A structural equation modeling methodology was used to determine how well the agile manufacturing model t the data and to identify support for each of the incorporated hypotheses. 3.1. Measurement of constructs The theorized model incorporates constructs related to JIT-production, JIT-purchasing, agile manufacturing, operational performance, and rm performance. The scales selected to measure the constructs are displayed in Appendix A. Agile manufacturers must exhibit capabilities of responsiveness, exibility, and quickness in responding to changes in customer demand (Shari and Zhang, 2001). The agile manufacturing scale was developed based on a prioritized listing of 20 capabilities necessary for organizations to achieve agility developed by Shari and Zhang (2001). A scale item was fashioned for each of Shari and Zhangs top 10 items. Respondents were asked to indicate their degree of agreement with each statement. Seven-point Likert scales were used with strongly disagree and strongly agree as anchors. JIT-production and JIT-purchasing focus on the elimination of waste and optimal utilization of resources in production and purchasing processes. JIT-production was measured using the multi-item scale developed by Brox and Fader (2002). Respondents were asked to indicate which of 13 JIT-production related practices had been implemented by their organizations. JIT-purchasing was measured with the 7-item scale developed by Germain and Drge (1997). Respondents were asked to indicate their degree of agreement with each statement. Seven-point Likert scales were used with strongly disagree and strongly agree as anchors. Operational performance was measured using a 13-item performance metrics scale developed by Bowersox et al. (2000). The items incorporate customer service, cost management, quality, productivity and asset management performance metrics. Respondents were asked to rate their organizations performance compared to that of their competitors on the operational performance metrics. The items were measured using 7-point Likert scales anchored with much worse than competition and much better than competition. Although Bowersox et al. (2000) used 5-point scales, the 7-point scales were adopted for consistency purposes. The scales for measuring the nancial and marketing performance of the rm were previously used by Green and Inman (2005) and Green et al. (2004). The nancial performance items were taken directly from Claycomb et al. (1999a). The marketing performance items were developed by Green and Inman (2005) based on measures of marketing performance (sales volume, market share and sales growth) identied by Kohli and Jaworski (1990). The items

in these scales were measured with 7-point Likert scales anchored with strongly disagree and strongly agree. 4. Results 4.1. Survey effectiveness A total of 1350 packets were mailed of which 18 were returned due to incorrect addresses. Further, 121 non-participating forms were returned. Ninety-six manufacturers responded with completed instruments for a response rate of 7.9%. This response rate is low but not atypical for industrial research. Other published works in similar circumstances yielded response rates as low as 7.5% (Nahm et al., 2003a,b), 6.7% (Tan et al., 2002), and 6.3% (Dwyer and Welsh, 1985). While Patterson et al. (2004) did not specically identify their response rate, they found it necessary to survey three different databases (one of higher-level managers and two of logistics managers) to gather only 107 responses. While manufacturing managers are the prime source for supply chain management related data, they are often under severe time and resource constraints making it difcult to achieve high response rates to surveys. Lambert and Harrington (1990, p. 21) describe a common approach to assessment as comparing the rst and second waves of responses and assuming that non-response bias is nonexistent if no differences exist on the survey variables. Following this common approach, respondents were categorized as responding to either the initial or follow-up requests sent approximately 2 weeks later. Those responding to the initial requests were classied as early responders; those responding to the follow-up requests were classied as late responders. Fifty-four percent (52) of the respondents were categorized as early respondents and 46% (44) were categorized as late respondents. A comparison of the means of the descriptive variables and the scale items for the two groups was conducted using one-way ANOVA. The comparisons resulted in statistically non-signicant differences at the .01 level. Because non-respondents have been found to descriptively resemble late respondents (Armstrong and Overton, 1977), this nding of general equality between early and late respondents indicates that non-response bias has not negatively impacted the assembled data set. When data for the independent and dependent variables are collected from single informants, common method bias may lead to inated estimates of the relationships between the variables (Podsakoff and Organ, 1986). As Podsakoff and Organ (1986) recommended, Harmans one-factor test was used post hoc to examine the extent of the potential bias. As prescribed by Harmans test, all variables were entered into a principal components factor analysis. According to Podsakoff and Organ (1986), substantial common method variance is signaled by the emergence of either a single factor or one general factor that explains a majority of the total variance. Results of the factor analysis revealed seven factors with eigenvalues greater than one, which combined to account for 70% of the total variance. While the rst factor accounted for 30% of the total variance, it did not account for a majority of the variance. Based upon these results of Harmans one-factor test, problems associated with common method bias are not considered signicant (Podsakoff and Organ, 1986). 4.2. Sample description All of the respondents indicated that they worked for manufacturing organizations. Seventy-two percent of the respondents identied themselves specically as plant or operations managers. The remaining 28% held management positions related to manufacturing, purchasing and distribution. Respondents averaged 5.5

348 Table 2 Scale assessment.

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

Reliability coefcients Scale Agile manufacturing JIT-production JIT-purchasing Operational performance Marketing performance Financial performance
*

GFI .937 .982 .971

* * * *

RMSEA .079 .087 .029

NNFI .973 .974 .991

* *

CFI .982 .991 .995

* *

Alpha .85 .78 .79 .80 .91 .92

Construct-reliability .89 .82 .53 .80 .95 .93

Variance-extracted .51 .54 .42 .58 .86 .77

NFI .956 .979 .966

* *

.996

.000

1.000

1.000

.998

Values not available for scales containing only 3 items.

years in their current positions. Mean sales revenues for the rms included in the sample were $7.7 billion, and the mean number of employees per rm was 21,211. Seventeen specic manufacturing SIC codes were identied. The most frequently identied SIC codes were: 34-fabricated metal products at 15.6%, 36-electronic and other electrical equipment at 7.3%, and 20-food and kindred products at 7.3%. Respondents represented 30 different states. The most frequently identied states were Ohio (10.4%), Michigan (9.4%), and Illinois (7.3%).

4.3. Scale assessment process Quality measurement scales must exhibit content validity, unidimensionality, reliability, discriminant validity, and convergent validity. Since all scales were taken directly from prior research (Shari and Zhang, 2001; Claycomb et al., 1999a,b; Brox and Fader, 2002; Green and Inman, 2005; Bowersox et al., 2000), content validity is assumed. Survey results used to assess all scales are found in Table 2. With the exception of operational performance and JITproduction, all scales were treated as rst-order factors (Garver and Mentzer, 1999). Bowersox et al. (2000) described operational performance as comprised of ve distinct factors: customer service, cost management, quality, productivity and asset management. To assess unidimensionality, operational performance was, therefore, treated as a second-order construct. Values for each of the ve factors were calculated by averaging across factor items, and the factor values were used in the unidimensionality assessment. Because responses to the JIT-production scale items were categorical (either implemented or not implemented), it was necessary to compute four composite measures by summing across the individual items in a manner similar to that recommended by Garver and Mentzer (1999). The original scale includes 17 items. KR20 reliability analysis indicated that the removal of item 7 (Preventive Maintenance Programs) would improve the overall reliability of the scale. The remaining 16 items were segmented into four groups to facilitate computation of the composites. Unidimensionality is indicated by goodness-of-t index (GFI) values greater than .90 (Ahire et al., 1996), non-normed-t index (NNFI) and comparative-t index (CFI) values greater than .90 (Garver and Mentzer, 1999), and root mean square error of approximation (RMSEA) below .08 (Garver and Mentzer, 1999). In order to achieve unidimensionality, it was necessary to remove items 6, 8 and 10 from the agile manufacturing scale and item 7 from the JIT-purchasing scale, and the cost management and asset management factors from the operational performance scale. After re-specication the agile manufacturing, JIT-production, JITpurchasing, and nancial performance scales all met the GFI, NNFI, and CFI minimums indicating unidimensionality. Because the operational performance and nancial performance scales contain only three items, it is not possible to compute GFI, NNFI, CFI, and RMSEA values. Principal components analysis, however, indicated that each scale measured only one dimension. The parameter estimates

for all scales were signicant and greater than .60 also indicating unidimensionality. The RMSEA values for the JIT-purchasing, agile manufacturing, and nancial performance were below the recommended .08 level. The RMSEA for JIT-production only slightly exceeded the .08 level at .087. Garver and Mentzer (1999) recommend computing Cronbachs coefcient alpha and the SEM construct-reliability and varianceextracted measures to assess scale reliability. They indicate that alpha and construct-reliability values greater than or equal to .70 and a variance-extracted measure of .50 or greater indicate sufcient reliability. Two of the three reliability coefcients for the JIT-purchasing scale exceed the recommended minimums (at .42, the variance-extracted measure was slightly below the desired .5). All other scales exceed the minimum reliability requirements on all three measures. Ahire et al. (1996) recommend assessing convergent validity using the normed-t index (NFI) coefcient with values greater than .9 indicating strong validity. Garver and Mentzer (1999) recommend reviewing the magnitude of the parameter estimates for the individual measurement items to assess convergent validity. A strong condition of validity is indicated when the estimates are statistically signicant and greater than or equal to .70. A weak condition of validity is indicated when estimates are statistically signicant but have values less than .70. While the NFI was not available for the operational performance and marketing performance scales, signicant parameter estimates greater than .70 indicate convergent validity for both. An NFI exceeding .95 and statistical signicance of the parameter estimates indicates sufcient convergent validity for all other scales. Gerbing and Anderson (1988) recommend that scales be tested for discriminant validity using a chi-square difference test for each pair of scales under consideration. A statistically signicant difference in chi-squares indicates discriminant validity (Garver and Mentzer, 1999; Ahire et al., 1996; Gerbing and Anderson, 1988). The Chi-square difference tests for pairings of each scale with other study scales returned signicant differences at the .01 level, indicating discriminant validity for all scales.

4.4. Measurement model Fig. 2 displays the measurement model that incorporates the scales described and assessed in the preceding section. As Koufteros (1999) recommends, the scales are further assessed within the context of the full measurement model using a conrmatory factor analysis methodology. The measurement model ts the data relatively well with a relative chi-square value of .95, an RMSEA value of 0.00, a CFI value of .99, and an NNFI value of 0.99. A review of the standardized residual matrix identied only four pairs with absolute values greater than 2.58 (OPA and MP5, JITPRB and FP4, JITPRB and MP6, and JITPRB and MP7). Looking at the individual items, we chose not to re-specify based on the importance of each item to the affected scales.

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

349

Fig. 2. Measurement model with standardized estimates and (t-values).

Relative chi-square = .95; Chi-square P-value = 0.72; RMSEA = 0.00; CFI = 0.99; NNFI = 0.99.

4.5. Structural equation modeling results Summary values for the study variables were computed by averaging across the items in the re-specied scales with the exception of the JIT-production scale for which items were summed. Descriptive statistics and the correlation matrix for the summary variables are presented in Table 3. Correlation coefcients are positive and signicant at the .05 level for all of the hypothesized relationships in the agile manufacturing model with the exception of the coefcients for JIT-purchasing and nancial performance, and JITproduction and operational performance. Fig. 1 depicts the theorized agile manufacturing performance model as structurally assessed. Fig. 3 illustrates the model with the

structural equation modeling results specied in the LISREL 8.7 output. Results relating to t of the model generally support a claim of good t. The relative chi-square (chi-square/degrees of freedom) value of 1.08 is less than the 3.00 maximum recommended by Kline (1998) and the root mean square error of approximation (.03) is below the recommended maximum of .08 (Schumacker and Lomax, 1996). The P-value associated with the chi-square is .17, above the recommended minimum of.05 (Byrne, 1998). Results associated with the t indices are somewhat mixed. The GFI (.79) and NFI (.88) are below the .90 level recommended by Byrne (1998). These indices are more heavily impacted by a relatively small sample size and, as Byrne (1998) points out, the Comparative-Fit Index (CFI) and Incremental-Fit Index (IFI) are more appropriate when

350 Table 3 Descriptive statistics and correlations. Summary variable A. Descriptive statistics (n = 96) Agilemanufacturing(AM) JIT-purchasing (JITPU) JIT-production (JITPR) Operational performance (OP) Financial performance (FP) Marketing performance (MP) AM B. Correlation matrix (n = 96) JITPU JITPR OP FP MP
* **

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

Mean 4.89 4.61 9.42 5.26 4.69 4.51 JITPU

Standard deviation .92 1.14 3.79 .69 1.20 1.23 JITPR OP FP

.484** .322** .477** .421** .377**

.531** .316** .134 .224*

.145 .206* .251*

.321** .386**

.640**

Correlation is signicant at the 0.05 level (2-tailed). Correlation is signicant at the 0.01 level (2-tailed).

the sample size is small. The CFI (.98) and IFI (.98) both exceed the recommended .90 level (Byrne, 1998). Of the ve study hypotheses, only the relationship between JITproduction and agile manufacturing is not supported by the results. The JIT-purchasing to agile manufacturing link (H1) is positive and signicant with an estimate of .65 and t-value of 3.17. The estimate of .10 for the link from JIT-production to agile manufacturing (H2) is non-signicant with a t-value of .56. The link from agile manufacturing to nancial performance (H3) is positive and significant with a standardized estimate of .49 and an associated t-value of 3.72. The agile manufacturing to marketing performance link (H4) is positive and signicant with a standardized estimate of .46 and associated t-value of 3.67. Finally, the agile manufacturing to operational performance link (H5) is positive and signicant with a standardized estimate of 0.58 and t-value of 4.13. The lack of support for the hypothesized link between JITproduction and agile-manufacturing is surprising and troubling. This result, combined with the modication indices, led us to rethink the model. This change in thought, coupled with Hair et al. (1998) recommendation for a competing models approach

to structural equation modeling when alternative formulations are suggested by underlying theory, prompted us to remove the path between JIT-production and agile manufacturing, making JITproduction antecedent to JIT-purchasing, thereby, indicating an indirect (mediation), rather than direct, link to agile manufacturing. While material purchase obviously must occur before production, most purchasing is based on production plans that anticipate scheduling and sequencing activities. Additionally, a review of the modication indices, resulting from assessment of the theorized model, suggests that an additional path from marketing performance to nancial performance be added. Inclusion of the additional path is supported by the results of a study by Green et al. (2006) which reported a positive relationship between marketing performance and nancial performance. Vazquez-Bustelo et al. (2007) found that the adoption of agile manufacturing positively impacts manufacturing strength thus leading to improved business performance, hence operational performance was treated as an antecedent to rm performance, i.e., marketing and nancial, performance. This alternative model and associated structural equation modeling results are illustrated in Fig. 4.

Fig. 3. Agile manufacturing hypothesized structural model with standardized estimates (** signicant at 0.01 level). Relative chi-square = 1.08; Chi-square P-value = 0.17; RMSEA = 0.03; CFI = 0.98; NNFI = 0.97.

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

351

Fig. 4. Agile manufacturing good-t structural model with standardized estimates (** signicant at 0.01 level). Relative chi-square = .96; Chi square P-value = 0.70; RMSEA = 0.00; CFI = 0.99; NNFI = 0.99.

The standardized estimate for the JIT-production to JITpurchasing link is .69 with an associated t-value of 4.88 (signicant at the .01 level). The link from JIT-purchasing to agile manufacturing remains positive and signicant. Rather than a direct positive relationship between JIT-production and agile manufacturing, based on this reformulation of the model, it appears that JIT-purchasing mediates the relationship between JIT-production and agile-manufacturing. The link from agile-manufacturing to operational performance remains positive and signicant. The estimate for the marketing to nancial performance link is 0.64 with a t-value of 4.96. Operational performance directly impacts marketing performance with a standardized estimate of .44 and t-value of 3.49. Operational performance does not directly impact nancial performance, however, with an estimate of .11 and t-value of 1.01. The impact of operational performance on nancial performance is mediated by marketing performance. The overall t improved with a relative chi-square = .96, a RMSEA of 0.00, a P-value of .70, a CFI of .99 and an NNFI of .99. The GFI (.81) and NFI (.89), however, remained below the desired .90 level. There was concern that environmental uncertainty may moderate the hypothesized relationship between agile manufacturing and operational performance. Following the general methodology described by Baron and Kenny (1986), moderation was assessed. However, the results indicated that environmental uncertainty did not moderate the relationship between agile manufacturing and organizational performance. Details of the analysis are found in Appendix B. 5. Discussion A broad sample of large U.S. manufacturers provided data for assessing the agile manufacturing performance model. Although some re-specication was necessary, all study scales were determined to be unidimensional, reliable, and valid. Results of the structural equation modeling analysis showed that the overall model t the data well and specically support all but one of the study hypotheses. The resulting support for the idea that JIT-purchasing is antecedent to agile manufacturing is not surprising. Within the manufacturing sector increased use of JIT-purchasing practices lead to improved agile manufacturing capabilities. This partially supports the theoretical literature that purports that leanness, specically JIT implementation, is a foundation or a precur-

sor to agility (mutually supportive) and the empirical ndings of Narasimhan et al. (2006) that when viewed from a performance [capability] perspective, leanness is a precursor to agility. Surprisingly only one of the two primary elements of JIT was found to support agility. The relationship between JIT-production and agile manufacturing was non-signicant. This result would seem to indicate that, within the context of the model, JIT-purchasing alone, rather than in combination with JIT-production, explains a signicant portion of the variation in agile manufacturing. This is inconsistent with the belief that JIT ow and other production related activities are precursors to agility. However, Narasimhan et al. (2006) note that other studies have shown that JIT ow is less signicant than other elements. Our nding does not support H2. The results for JIT-production do not support the notion that JIT-production is antecedent to agile manufacturing nor does it support the notion that the two are mutually exclusive. This may be in agreement with McCullen and Towills (2001) argument that agile manufacturing can subsume the paradigm of lean production. Is the production aspect of JIT so much a part of agility that one may not distinguish a difference between JIT-production and the production element within an agile manufacturing rm? Narasimhan et al. (2006) note that agile does imply that many of the principles and techniques of lean manufacturing are in place. If the JIT-production element is already in place then increased supplier/customer integration, in the form of high levels of JIT-purchasing, could show a far greater impact on agility than JIT-production alone. This is one possible explanation for the lack of support for Hypothesis 2 in the original model. The move from JIT to agile could involve keeping the JIT-production element of lean constant but greatly increasing the emphasis on JIT-purchasing. Although no such move is tested directly, the conceptual argument is consistent with the results of the alternate model in which JIT-production is assessed as antecedent to JIT-purchasing. The idea that manufacturing excellence may generate a certain level of performance, but that additional improvement requires the level of supply chain integration suggested by JIT-purchasing is consistent with outward-facing rms in Frohlich and Westbrooks (2001) classication of rms based on arcs of integration. In addition it was found that organizations that become agile manufacturers can expect improved operational and rm performance. This nding is consistent with Vazquez-Bustelo et al. (2007) who found that the adoption of agile manufacturing positively impacts manufacturing strength thus leading to improved

352

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

business performance and Narasimhan et al. (2006) who found that agile rms exceed lean and other rms on most performance measures used. Interestingly, this is consistent with the proposed relationships among three of the four perspectives in balanced scorecard logic; customer perspective (marketing performance), internal business perspective (operational performance), and nancial perspective (nancial performance). In summary, we offer the following proposal: In the manufacturing sector, JIT-purchasing combined with JITproduction enhances a rms manufacturing agility. Improved manufacturing agility leads to the improved operating performance of the rm, which in turn leads to the improved marketing and nancial performance of the rm. 5.1. Limitations of the study While the objectives of the study were successfully accomplished, limitations of the study should be noted. The response rate raised concerns of potential non-response bias. Although the two waves of responses were compared and no evidence of bias was noted, a more direct assessment of the potential bias utilizing data from a third wave and an intensive follow-up on non-respondents may have strengthened the study. Because responses related to both the dependent and independent variables were collected from the same individual, the potential for common method bias was also a concern. While subsequent testing for the bias relieved the concern, collection of the strategy and performance data from separate sources would also have strengthened the study. Also, since all measures were at the organization level, not the individual plant level, data from multi-plant rms could dilute the data if some plants were focused on lean and others on agility. However, this should weaken the results rather than articially strengthen them. There is concern that the measurement scales conceptualizing the JIT-related constructs were borrowed from different research streams and that the formats and structuring of the scales is inconsistent. The JIT-purchasing scale is Likert-based, requiring respondents to indicate degree of agreement. The JIT-production scale is categorically structured requiring respondents to indicate whether or not their organizations have adopted a particular JITproduction practice. In future research efforts, we recommend that the scales be reformulated for consistency. The study focused on large U.S. manufacturers because we felt this group is more likely to have adopted JIT and agile practices. As a result, it may not be appropriate to generalize results to medium and small manufacturers. Further, the theory as developed and tested applies only in the manufacturing sector. Caution should be exercised when generalizing the results to the service and governmental sectors. 5.2. Future research This study links JIT practices to manufacturing agility and agility to performance. Additional research aimed at verifying these results is necessary. Sample frames that focus on small and medium-sized manufacturers are necessary to facilitate generalization of these results. It may also be advantageous to view the combination of JIT-purchasing (with JIT-production as antecedent) and agile manufacturing as an overall supply chain strategy. Also, further research could incorporate the other elements of lean manufacturing such as TQM, preventive maintenance and human resource management. Once the impact of each element has been evaluated comparisons can be made between the effects of individual elements compared to the effect of all elements working synergistically [Shah and Wards (2003) term applied

to the four bundles of lean manufacturing working in concert]. Although uncertainty was not included in our models, concern that it may moderate the hypothesized relationship between agile manufacturing and operational performance led to subsequent testing for moderation external to our models. Results indicated that environmental uncertainty did not moderate the relationship between agile manufacturing and organizational performance supporting our original determination not to include it in the analysis. Details of the assessment for moderation are presented in Appendix B. While environmental uncertainty did not play a signicant role in our models (see Appendix B), future studies could be expanded by including market environment (degree of turbulence or degree of uncertainty) as a variable. It would be informative to examine the degree of match between environments (stable/lean vs. turbulent/agile) and determine how degree of match impacts operational, nancial and marketing performance. Finally, the study could be strengthened by the inclusion of items that determine if the home plant of the respondent, usually a plant manager, is one of multiple plants in an organization. This knowledge may lead to further interesting analysis of multi-plant rms where some plants are focused on lean and others on agility.

Appendix A. Measurement scales Agile manufacturing (alpha = .85) Note: Items 6, 8 and 10 removed to achieve unidimensionality. Please indicate the extent to which you agree or disagree with each statement. (1 = strongly disagree, 7 = strongly agree) 1. This organization has the capabilities necessary to sense, perceive and anticipate market changes. 2. The production processes of this organization are exible in terms of product models and congurations. 3. This organization reacts immediately to incorporate changes into its manufacturing processes and systems. 4. This organization has the appropriate technology and technological capabilities to quickly respond to changes in customer demand. 5. This organizations strategic vision emphasizes the need for exibility and agility to respond to market changes. 6. This organization has formed co-operative relationships with customers and suppliers. 7. This organizations managers have the knowledge and skills necessary to manage change. 8. This organization has the capabilities to meet and exceed the levels of product quality demanded by its customers. 9. This organization has the capabilities to deliver products to customers in a timely manner and to quickly respond to changes in deliver requirements. 10. This organization can quickly get new products to market. JIT-Purchasing (alpha = .79) Note: Item 7 removed to achieve unidimensionality. Please indicate the extent to which agree or disagree with each statement (1 = strongly disagree, 7 = strongly agree). 1. 2. 3. 4. 5. 6. 7. Orders are placed to suppliers and delivered on a daily basis. Our suppliers warehouses/factories are located nearby. Production plans are shared with suppliers. Small lot size orders are placed with suppliers. Inspection of incoming materials has been reduced. Our staff visits suppliers plants on an informal basis. We involve suppliers in new product/materials design.

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355

353

JIT-Production (KR20 = .818, alpha based on composites = .78) Item 7 was removed based on KR20 assessment. Please indicate which of the following JIT practices have been implemented in your organizations production processes. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Kanban Integrated product design Integrated supplier network Plan to reduce setup time Quality circles Focused factory Preventive maintenance Line balancing Education about JIT Level schedules Stable cycle rates Market-paced nal assembly Group technology Program to improve quality (Product) Program to improve quality (Process) Fast inventory transportation system Flexibility of workers skill

12. Inventory turn 13. Return on assets Financial Performance (alpha = .92) Please rate your organizations performance in each of the following areas as compared to the industry average. (1 = well below industry average; 7 = well above industry average) 1. 2. 3. 4. Average return on investment over the past 3 years. Average prot over the past 3 years. Prot growth over the past 3 years. Average return on sales over the past 3 years.

Marketing Performance (alpha = .91) Please rate your organizations performance in each of the following areas as compared to the industry average. (1 = well below industry average; 7 = well above industry average) 1. Average market share growth over the past 3 years. 2. Average sales volume growth over the past 3 years. 3. Average sales (in dollars) growth over the past 3 years. Appendix B. Moderating impact of environmental uncertainty Following the general methodology described by Baron and Kenny (1986), moderation is assessed. When testing for moderation, it is desirable that the moderator variable (EU) be uncorrelated with the predictor (AM) and criterion (OP) variables (Baron and Kenny, 1986). Descriptive statics are displayed in Table B1. The
Table B1 Descriptive statistics. Mean Agile manufacturing Operational performance Environmental uncertainty 4.89 5.26 3.61 Standard deviation .92 .69 1.06

Operational performance (alpha = .80) Note: The cost management and asset management factors were removed to achieve unidimensionality. Please rate your companys performance in each of the following areas as compared to the performance of your competitors. (1 = much worse than competition, 7 = much better than competition) Customer service 1. Customer satisfaction 2. Product customization 3. Delivery speed Cost management 4. Logistics cost Quality 5. Delivery dependability 6. Responsiveness 7. Order exibility 8. Delivery exibility Productivity 9. Information systems support 10. Order ll capacity 11. Advance ship notication Asset management
Table B3 Coefcients for agile manufacturing, environmental uncertainty, and interaction. Model Unstandardized coefcients B 1 (Constant) AM EU AM EU 2.591 .496 .262 .040 Std. error 1.119 .216 .311 .060

Table B2 Correlations. AM Agile manufacturing (AM) Operational performance (OP) Environmental uncertainty (EU)
**

OP .477** 1 .105

EU .038 .105 1

1 .477** .038

Signicant at the 0.01 level (2-tailed).

Standardized coefcients Beta .656 .399 .372

t-value

Signicant

Collinearity statistics Tolerance VIF 9.871 27.096 37.111

2.315 2.296 .844 .672

.023 .024 .401 .503

.101 .037 .027

Dependent variable: op.

Table B4 Coefcients for agile manufacturing and interaction. Model Unstandardized coefcients B 1 (Constant) AM AM EU 3.490 .327 .009 Std. Error .342 .080 .011 Standardized coefcients Beta .432 .086 10.202 4.065 .812 .000 .000 .419 t-value Signicance Collinearity statistics Tolerance .729 .729 VIF 1.372 1.372

Dependent variable: op.

354

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355 Green Jr., K.W., McGaughey, R., Casey, K.M., 2006. Does supply chain management strategy mediate the association between market orientation and organizational performance? Supply Chain Management: An International Journal 11 (5), 407414. Green Jr., K.W., Medlin, B., Whitten, D., 2004. Developing optimism to improve performance: an approach for the manufacturing sector. Industrial Management & Data Systems 104 (2), 106114. Gunasekaran, A., 1999a. Design and implementation of agile manufacturing systems: editorial. International Journal of Production Economics 62, 16. Gunasekaran, A., 1999b. Agile manufacturing: a framework for research and development. International Journal of Production Economics 62, 87105. Gunasekaran, A., Lai, K.-H., Cheng, T.C.E., 2008. Responsive supply chain: a competitive strategy in a networked economy. Omega: The International Journal of Management Science 36, 549564. Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate Data Analysis, 5th ed. Prentice Hall, New Jersey. Harrison, A., 1997. From leanness to agility. Manufacturing Engineer 76 (6), 257260. Hormozi, A.M., 2001. Agile manufacturing: the next logical step. Benchmarking: An International Journal 8 (2), 132143. Jin-Hai, Li, Anderson, A.R., Harrison, R.T., 2003. The evolution of agile manufacturing. Business Process Management 2 (9), 170189. Katayama, H., Bennett, D., 1999. Agility, adaptability and leanness: a comparison of concepts and a study of practice. International Journal of Production Economics 6061, 4351. Kidd, P.T., 1994. Agile Manufacturing: Forging New Frontiers. AddisonWesley, Reading, MA and Workingham. Kline, R.B., 1998. Principles and Practice of Structural Equation Modeling. Guilford Press, New York. Kohli, A.K., Jaworski, B.J., 1990. Market orientation: the construct, research propositions, and managerial implications. Journal of Marketing 54 (2), 118. Koufteros, X.A., 1999. Testing a model of pull production: a paradigm for manufacturing research using structural equation modeling. Journal of Operations Management 17 (4), 467488. Krishnamurthy, R., Yauch, C.A., 2007. Leagile manufacturing: a proposed corporate infrastructure. International Journal of Operations and Production Management 27 (6), 588604. Lambert, D.M., Harrington, T.C., 1990. Measuring nonresponse bias in customer service mail surveys. Journal of Business Logistics 11 (2), 525. Maskell, B., 2001. The age of agile manufacturing. Supply Chain Management 6 (1), 511. Mason-Jones, R., Naylor, B., Towill, D.R., 2000. Lean, agile or leagile? Matching your supply chain to the marketplace. International Journal of Production Research 38 (17), 40614070. McCullen, P., Towill, D., 2001. Achieving lean supply through agile manufacturing. Integrated Manufacturing Systems 12 (7), 524533. Mehra, S., Inman, R.A., 1992. Determining the critical elements of JIT implementation. Decision Sciences 23 (1), 160174. Miller, D., Drge, C., 1986. Psychological and traditional determinants of structure. Administrative Science Quarterly 31 (4), 539560. Nahm, A.Y., Vonderembse, M.A., Koufteros, X.A., 2003a. The impact of organizational culture on time-based manufacturing and performance. Decision Sciences 35 (4), 579607. Nahm, A.Y., Vonderembse, M.A., Koufteros, X.A., 2003b. The impact of organizational structure on time-based manufacturing and plant performance. Journal of Operations Management 21 (3), 281306. Narasimhan, R., Swink, M., Kim, S.W., 2006. Disentangling leanness and agility: an empirical investigation. Journal of Operations Management 24, 440457. Naylor, J.B., Naim, M.M., Berry, D., 1999. Leagility: integrating the lean and agile manufacturing paradigms in the total supply chain. International Journal of Production Economics 62, 107118. Patterson, K.A., Grimm, C.M., Corsi, T.M., 2004. Diffusion of supply chain technologies. Transportation Journal 43 (3), 523. Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research: problems and prospects. Journal of Management 12 (4), 531544. Ramesh, G., Devadasan, S.R., 2007. Literature review on the agile manufacturing criteria. Journal of Manufacturing Technology Management 18 (2), 182 201. Richards, C.W., 1996. Agile manufacturing: beyond lean. Production and Inventory Management Journal 37 (2), 6064. Robertson, M., Jones, C., 1999. Application of lean production and agile manufacturing concepts in a telecommunications environment. International Journal of Agile Manufacturing Systems 1 (1), 1416. Sarkis, J., 2001. Benchmarking for agility. Benchmarking: An International Journal 8 (2), 88107. Schumacker, R.E., Lomax, R.G., 1996. A Beginners Guide to Structural Equation Modeling. Lawrence Erlbaum Associates, Inc., New Jersey. Shah, R., Ward, P.T., 2007. Dening and developing measures of lean production. Journal of Operations Management 25, 785805. Shah, R., Ward, P.T., 2003. Lean manufacturing: context, practice bundles, and performance. Journal of Operations Management 21, 129149. Shari, H., Zhang, Z., 2001. Agile manufacturing in practice: application of a methodology. International Journal of Operations & Production Management 21 (56), 772794. Tan, K.C., Lyman, S.B., Wisner, J.D., 2002. Supply chain management: a strategic perspective. International Journal of Operations and Production Management 22 (6), 614631.

Table B5 Environmental uncertainty adapted from Miller and Drge (1986). Please indicate the extent to which you agree or disagree with each statement. (1 = strongly disagree, 7 = strongly agree) 1. This organization must change its marketing practices frequently 2. The actions of this organizations competitors are unpredictable 3. The demands and tastes of this organizations customers are almost unpredictable 4. It is necessary to frequently make major changes in this organizations production processes 5. This organizations products become obsolete at a rapid rate

correlations are presented in Table B2 below indicate that EU is not signicantly correlated with either AM or OP. Moderation is supported if the interaction (XY) is signicant (Baron and Kenny, 1986). The results of regressing AM, EU, and XY against OP are presented in Table B3. While the regression coefcient for the interaction (XY) is not signicant, it should be noted that multicolinearity is present making the coefcients difcult to interpret. Table B4 displays the results of regressing AM and XY against OP without EU present in the model. Multicolinearity is not present. The regression coefcient for XY is .009 with an associated t-value of .812 and a computed signicance level of .419. Based on these results, it is concluded that EU does not moderate the relationship between AM and OP. EU is measured using a 5-item scale adapted from Miller and Drge (1986) displayed in Table B5. References
Ahire, S.L., Golhar, D.Y., Waller, M.A., 1996. Development and validation of TQM implementation constructs. Decision Sciences 27 (1), 2356. Armstrong, J.S., Overton, T.S., 1977. Estimating nonresponse bias in mail surveys. Journal of Marketing Research 14 (3), 396402. Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology 51 (6), 11731182. Booth, R., 1996. Agile manufacturing. Engineering Management Journal 6 (2), 105112. Bowersox, D.J., Closs, D.J., Stank, T.P., Keller, S.B., 2000. How supply chain competency leads to business success. Supply Chain Management Review 4 (4), 7078. Brox, J.A., Fader, C., 2002. The set of just-in-time management strategies: an assessment of their impact on plant-level productivity and input-factor substitutability using variable cost function estimates. International Journal of Production Research 40 (12), 27052720. Byrne, B.M., 1998. Structural Equation Modeling with LISREL, PRELIS, and SIMPLIS. Lawrence Erlbaum Associates, New Jersey. Christopher, M., Towill, D., 2001. An integrated model for the design of agile supply chains. International Journal of Physical Distribution & Logistics Management 31 (4), 235246. Claycomb, C., Drge, C., Germain, R., 1999a. The effect of just-in-time with customers on organizational design and performance. International Journal of Logistics Management 10 (1), 3758. Claycomb, C., Germain, R., Drge, C., 1999b. Total system JIT outcomes: inventory, organization and nancial effects. International Journal of Physical Distribution and Logistics 29 (10), 612630. Dwyer, F.R., Welsh, M.A., 1985. Environmental relationships of the internal political economy of marketing channels. Journal of Marketing Research 22 (4), 397414. Freeland, J.R., 1991. A survey of just-in-time purchasing practices in the United States. Production and Inventory Management Journal 32 (2), 4349. Frohlich, M.T., Westbrook, R., 2001. Arcs of integration: an international study of supply chain strategies. Journal of Operations Management 19, 185200. Garver, M.S., Mentzer, J.T., 1999. Logistics research methods: employing structural equation modeling to test for construct validity. Journal of Business Logistics 20 (1), 3357. Gerbing, D.W., Anderson, J.C., 1988. An updated paradigm for scale development incorporating unidimensionality and its assessment. Journal of Marketing Research 25 (2), 186192. Germain, R., Drge, C., 1997. Effect of just-in-time purchasing relationships on organizational design, purchasing department conguration, and rm performance. Industrial Marketing Management 26 (2), 115125. Goldman, S.L., Nagel, R.N., 1993. Management, technology and agility: the emergence of a new era in manufacturing. International Journal of Technology Management 8 (12), 1838. Goldsby, T.J., Grifn, S.E., Roath, A.S., 2006. Modeling lean, agile, and leagile supply chain strategies. Journal of Business Logistics 27 (1), 5780. Green Jr., K.W., Inman, R.A., 2005. Using a just-in-time selling strategy to strengthen supply chain linkages. International Journal of Production Research 43 (16), 34373453.

R.A. Inman et al. / Journal of Operations Management 29 (2011) 343355 Vazquez-Bustelo, D., Avella, L., Fernandez, E., 2007. Agility drivers, enablers and outcomes. International Journal of Operations and Production Management 27 (12), 13031332. Wisner, J.D., Leong, G.K., Tan, K.-C., 2005. Principles of Supply Chain Management: A Balanced Approach. Southwestern, United States. Yusuf, Y.Y., Adeleye, E.O., 2002. A comparative study of lean and agile manufacturing with a related survey of current practices in the UK. International Journal of Production Research 40 (17), 45454562.

355

Zhang, Z., Shari, Hossein, 2007. Towards theory building in agile manufacturing strategy a taxonomical approach. IEEE Transactions on Engineering Management 54 (2), 351370. Zhang, Z., Shari, H., 2000. A methodology for achieving agility in manufacturing organizations. International Journal of Operations and Production Management 20 (4), 496512.