Effects of simplicity and discipline on operational flexibility: An empirical reexamination of the rigid flexibility model

Journal of Operations Management 24 (2006) 932–947 www.elsevier.com/locate/jom

Effects of simplicity and discipline on operational flexibility: An empirical reexamination of the rigid flexibility model Giovani J.C. da Silveira * Haskayne School of Business, University of Calgary, 2500 University Drive NW, Calgary, Alta., Canada T2N 1N4 Received 23 June 2003; received in revised form 28 December 2004; accepted 30 September 2005 Available online 15 December 2005

Abstract This study empirically tests the rigid flexibility model proposed by Collins and Schmenner [Collins, R.S., Schmenner, R.W., 1993. Achieving rigid flexibility: factory focus for the 1990s. European Management Journal 11 (4), 443–447]. It investigates relationships between flexibility performance and adoption of simplicity and discipline programs in manufacturing. The research replicates the study by Collins et al. [Collins, R.S., Cordon, C., Julien, D., 1998. An empirical test of the rigid flexibility model. Journal of Operations Management 16 (2–3), 133–146] with some modifications, including the use of a broader international database, the assessment of both technology and organizational programs, and the testing of the moderating role of dedicated line layout on the relationships between simplicity, discipline and flexibility. Analysis of data from 285 manufacturers of fabricated metal products, machinery, and equipment from 14 countries indicates that simplicity and discipline related positively to performance in product customization, volume flexibility, mix flexibility, and time to market, and that some of these relationships were more positive in high volume processes than in low volume processes. The results provide empirical validation to the rigid flexibility model in an international manufacturing context. # 2005 Elsevier B.V. All rights reserved. Keywords: Flexibility; Operations improvement; Regression analysis; Empirical research

1. Introduction Building flexibility to respond quickly to changing market needs has been regarded as one of the major challenges in operations management over recent years (Bordoloi et al., 1999; Barnes-Schuster et al., 2002). Industrial markets have been increasingly subject to frequent changes regarding product variety and demand volumes (Bayus and Putsis, 1999; Jack and Raturi, 2002). In most cases, however, customers do not accept paying higher prices or waiting longer for products

* Tel.: +1 403 220 6975; fax: +1 403 210 3327. E-mail address: [email protected]. 0272-6963/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jom.2005.11.004

fitting to new demands. For many operations, the challenge is how to build flexibility at no expense to cost, quality, or delivery performance (Boynton et al., 1993; Suarez et al., 1996). Over the last decade, a great deal of research has aimed at tackling such flexibility challenge (De Toni and Tonchia, 1998). Most authors focused on either exploring the relationship between flexibility and performance, or building conceptual typologies or taxonomies (Narasimhan and Das, 1999). However, few studies focused on the links between flexibility and operations improvements. Among those, Collins and Schmenner’s (1993) rigid flexibility model appears to provide one of the most consistent answers to producers squeezed by market volatility.

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

The rigid flexibility model suggested that flexibility competence could be developed by building simplicity and discipline in manufacturing. Simplicity was about streamlining information and materials flow processes. Discipline was about carrying out procedures in dedicated and consistent fashion. Both simplicity and discipline would result from improvements in several areas including information and process technology, labor development, product design, and process configuration. The model’s premise was somewhat paradoxical, as flexibility would result not from building capacity or inventory buffers [as suggested by several studies in operations and supply chain management, e.g. Fisher, 1997; Huang et al., 2002; Jack and Raturi, 2002] or from allowing improvisation in manufacturing. Instead, flexibility would result from rigid processes that consistently and diligently pursued strategic tasks: ‘‘. . . if the requirement is flexibility, then an atmosphere of permissiveness cannot be tolerated’’ (Collins and Schmenner, 1993, p. 444). Simplicity, rather than reducing the number of options available to the firm, should provide a streamlined process that was easier to reconfigure and adapt to changing requirements. Discipline, rather than stiffening procedures and skills, should promote the best practices and work methods that enabled the firm to respond to market changes. Despite the model’s appeal and influence in the operations strategy field [providing foundations for studies on trade-offs and the world-class paradigm, e.g. Noble, 1995; Flynn and Flynn, 1999; Beach et al., 2000], there has been surprisingly limited research to validate its propositions. So far, only Collins et al. (1998) appear to have developed an empirical test. They provided evidence to relationships between simplicity/ discipline and flexibility in manufacturers from the five western European countries of Britain, Germany, Switzerland, the Netherlands, and Finland. No study appears to have tested the model by using a broader geographical base, cross-examining different process types, or assessing the role of the manufacturing and information technologies that today appear critical to flexibility performance. Thus, while Collins et al. (1998) provided a valuable contribution in validating the model in a specific context, more research is needed to assess its applicability in a broader framework. This study addresses that research requirement. It searches for evidence to the rigid flexibility model through using a broad international database, building scales for simplicity and discipline that incorporate both technology and organizational approaches, and exploring relationships in high and low volume processes. Furthermore, the analysis focuses on core flexibility

933

dimensions including product customization, volume flexibility, mix flexibility, and time to market. The study uses data on the flexibility performance and improvement programs of 285 manufacturers of fabricated metal products, machinery, and equipment from 14 countries. The research aims to replicate the study by Collins et al. (1998), while incorporating some modifications to provide further knowledge about the model’s applicability in different contexts. 2. Background 2.1. Manufacturing flexibility Flexibility of a system has been defined as its ability to adapt to environmental change (Sethi and Sethi, 1990; Gupta and Goyal, 1989). It entails modifying processes and product configurations with little penalty in time or cost to deal with changing circumstances (Slack, 1987; Upton, 1994; Van Dijk, 1995). According to Bordoloi et al. (1999) and BarnesSchuster et al. (2002), flexibility has grown in priority due to pressures to respond to changing market needs and shortening product life cycles. Several studies provided evidence to increasing volatility in customer demand. For example, a majority of respondents to ˚ hlstro¨m and Westbrook’s (1999) survey of British A operations managers indicated that customers were requiring increasing levels of customization and nonstandard products. Mendelson and Pillai (1999) found a steady decrease (average 9.4% per year) in the duration of product life cycles in several industry segments over the period from 1988 to 1995. Similar trends appear to affect requirements in product mix (Kekre and Srinivasan, 1990; Karuppan and Ganster, 2004), and demand volume (Suarez et al., 1996; Jack and Raturi, 2002). Research on flexibility is extensive and appears to have peaked in the 1990s. Sethi and Sethi (1990) and De Toni and Tonchia (1998) provided broad literature reviews. The major interest in research appears to be the classification of flexibility. Several authors have used organizational, hierarchical, temporal, or objective criteria to build flexibility taxonomies (De Toni and Tonchia, 1998). The most common taxonomies use objective criteria to define flexibility types corresponding to different elements of a production system. One may classify these in two groups. The first group involves taxonomies by authors such as Browne et al. (1984), Sethi and Sethi (1990), and Gerwin (1993). They provided comprehensive classifications involving types such as machine, materials, production, volume, and routing flexibility. Each flexibility type can be

934

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

defined by the ability to carry out changes in the corresponding object (e.g. machine, materials, production plans) with limited cost or time [pursuant to the rationale in Slack, 1983 that flexibility competencies are located in the two-dimensional space of range (the scope of states that can be assumed) and response (the time and cost to assume a new state)]. As an example, Gerwin (1993, p. 400) defined machine flexibility as, ‘‘. . . the types of operations performed [by the machine] without difficulty in switching from one to another’’. The second group of taxonomies has used a more aggregated perspective to flexibility. Slack (1987), Bartezzaghi and Turco (1989), Suarez et al. (1991), and Chen et al. (1992) coincided on three major types of flexibility at the system level, namely (i) volume flexibility (the ability to operate economically at different production volumes), (ii) mix flexibility (the ability to change the variety of products being made in a period), and (iii) product flexibility (the ability to design new products or modify existing ones). Additional dimensions in each study included ‘delivery flexibility’ in Slack (1987), ‘readiness’ in Bartezzaghi and Turco (1989), ‘delivery-time flexibility’ in Suarez et al. (1991), and ‘expansion flexibility’ in Chen et al. (1992). 2.2. Flexibility and performance Over the last two decades, several studies have provided evidence for the relationship between flexibility and performance in operations. Swamidass and Newell (1987) found a significant relationship between manufacturing flexibility and growth in sales and profitability in a sample of 35 companies. Kekre and Srinivasan (1990) found evidence in the profit impact of marketing strategies (PIMS) database that product line breadth was linked to performance in market share and return on investment. Fiegenbaum and Karnani (1991) suggested that output (volume) flexibility was associated to extra profit in small firms, especially in industries under strong demand fluctuation. Narasimhan and Das (1999) found a significant relationship between modification (product customization) flexibility and manufacturing cost reduction in a sample of 68 companies; however, non-significant relationships were found between cost, and volume and new product flexibility. Jack and Raturi (2002) found evidence to association between volume flexibility, and financial performance and delivery performance. Finally, Pagell and Krause’s (2004) replication of earlier studies by Swamidass and Newell (1987) and Pagell and Krause (1999) found evidence that increased flexibility led to improved performance, although this effect could not be

linked to the plants’ level of response to environmental uncertainty. 2.3. The rigid flexibility model Collins and Schmenner (1993) developed the rigid flexibility model to provide an answer for producers who felt squeezed by the need to meet increasing demands for flexibility at no expense to cost, quality, or delivery. The need for flexibility responsiveness came in requests such as changing product features, quickly developing new products, and frequently delivering small product volumes. They indicated that such flexibility competencies could be achieved through building simplicity and discipline in operations. The concepts of simplicity and discipline were more clearly defined in the empirical study by Collins et al. (1998). Simplicity was about streamlining information and material flow processes. It involved process focus, automation, and reduction in operations complexity. Discipline referred to reliability and organization in materials and information processing. It involved identifying and solving problems, improving work methods, and carrying out procedures in a dedicated and consistent fashion. Simplicity initiatives included product modularization, cellular layout, less wasted motion, inventory reduction, improved information exchange with suppliers, and greater visibility of stocks and material flows. Discipline initiatives included preventive maintenance, workplace development and housekeeping, continuous improvements, and operator checking of quality. Fig. 1 illustrates the constructs and relationships proposed by the rigid flexibility model. Collins et al. (1998) appear to have carried out the only empirical test of the model thus far. They merged two existing databases (Made in Europe and Made in Switzerland) to test associations between simplicity/ discipline and flexibility in 485 manufacturing plants from five European countries. They initially selected 18 items to assess flexibility, simplicity, and discipline in the database. Factor analyses were used to build the independent and dependent scales. They used six factors as surrogate measures of flexibility, namely inventory management, warehousing, product cycle time, total cycle time, new product introduction, and customer delivery time. They also generated two simplicity factors (lean management systems, lean attitude) and four discipline factors (waste, process capability, strategic processes, quality management processes). Hence, the analysis estimated the significance of 36 regression coefficients for relationships between the six simplicity/ discipline factors and the six flexibility factors. The

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

935

Fig. 1. The rigid flexibility model.

results suggested that 21 of those 36 coefficients were positive and significant ( p < .10). They also performed regression analyses for each country, with good support in the British and German samples, but limited support in the Swiss and Dutch samples (the Finnish sample was excluded from that analysis, probably due to a small number of observations). 2.4. Reexamining the rigid flexibility model Collins et al. (1998) provided empirical support to the rigid flexibility model in a database from five European countries. However, as discussed above, there is still need for more research to validate the model’s application in a broader context and using a modified set of variables. This study replicates the objectives and analytical procedures of the previous work. It similarly searches for evidence to positive relationships between simplicity/discipline and flexibility in manufacturing. The study also uses data from an international database focused on manufacturing strategies, practices, and performance. Independent variable scales are similarly constructed by selecting database variables that appear consistent with the initiatives that Collins and Schmenner (1993) associated to either simplicity or discipline. Finally, regression analysis is used to test the links between flexibility, and simplicity and discipline. However, since this study aims to test the model in a broader context, it introduces some innovations. Performance is measured by respondents’ perception to improvements in traditional flexibility dimensions,

namely product customization, volume flexibility, mix flexibility, and time to market. The objective here is to validate the model by using flexibility measures that may be more easily recognized than the flexibility factors used in the previous work, which included items such as inventory management, warehousing, and product cycle time. The second change concerns the attempt to validate the model in a broader international context. The International Manufacturing Strategy Survey (IMSS) gathers data from manufacturing companies from a larger variety of countries than the Made in Europe and Made in Switzerland databases. On the other hand, since IMSS is focused on fewer industrial sectors than the Made in Europe and Made in Switzerland databases, this study’s broader international scope will be counterbalanced by an industrial scope that is narrower than in the previous study. Finally, this study incorporates not only organizational initiatives but also technology-based initiatives into the simplicity and discipline scales. The previous study used mostly organizational items to build two simplicity and four discipline factors. The incorporation of technology items appears particularly relevant in light of studies that emphasized the potential of both technology and organizational methods to support flexibility. Adler (1985) discussed the potential of automation in design, manufacturing, and administration to leverage improvements in flexibility. Thilander’s (1992) study of the chemical industry indicated that companies handled environmental change through both technical means such as automatic control and

936

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

information transfer to operations, and organizational means such as multi-skilling and standardization of procedures. Gerwin (1993) suggested that a series of technologies and organizational methods were available to improve flexibility, but managers should know how to combine such methods to achieve superior results. Using a more integrative approach, Small and Chen (1997) indicated that organizational development strategies were critical for realizing the flexibility potential of advanced manufacturing technologies. These studies indicate that flexibility may benefit from combined investments in organization and technology programs. Thus, the present research aims to search for evidence to the rigid flexibility model in a broader framework. Besides replicating the objectives in Collins et al. (1998), it introduces new elements in research design. The research hypotheses follow the basic model thrusts with the flexibility dimensions discussed earlier: Hypothesis 1. The implementation of simplicity programs is positively related to product customization, volume flexibility, mix flexibility, and time to market in manufacturing. Hypothesis 2. The implementation of discipline programs is positively related to product customization, volume flexibility, mix flexibility, and time to market in manufacturing. 2.5. A theoretical extension A further aspect to be investigated is the extent to which the rigid flexibility model can be equally valid at high volume/low variety processes with a linear layout, and at low volume/high variety processes with a jobshop layout. There is some evidence to suggest that producers of low volume/high variety products might in fact lose flexibility if their processes were overly simple and rigid, as they often need to deploy significant workin-process and buffer capacity to cope with changing market requirements (Fisher, 1997; Huang et al., 2002). Moreover, several practices that Collins and Schmenner (1993) associated with the development of simplicity and discipline in operations, e.g. process focus, automation, and even-paced production planning are more suggestive of repetitive environments. Thus, the final hypothesis tests the role of dedicated lines as a moderator in relationships between flexibility and simplicity/ discipline. In specific, it investigates whether the development of simplicity and discipline programs in high volume plants using dedicated lines yields stronger flexibility improvements than their development in low volume plants using a job-shop or cellular layout:

Hypothesis 3. Use of dedicated lines moderates the relationships between simplicity/discipline programs and product customization, volume flexibility, mix flexibility, and time to market in manufacturing.

3. Methods 3.1. Sample Analysis was performed using data from the third round of the International Manufacturing Strategy Survey (IMSS-III). IMSS aims to identify the strategies, practices, and performance of manufacturing firms worldwide. It has been carried out since 1992 by an international network of operations strategy researchers. The third edition, released in July 2002, included responses from 14 countries. Survey companies were in ISIC Rev. 2 Division 38, Manufacture of Fabricated Metal Products, Machinery, and Equipment. The survey focuses on that division because of its large presence worldwide and pioneering role in implementing advanced operations technologies and practices. By focusing on a single industry, the survey aims to guarantee homogeneity and comparability of subject data in an international context. The survey was administered in each country by local research coordinators. Wherever necessary, the original questionnaire was translated from English by operations strategy academics. Questionnaires were mailed or emailed to the Director of Operations/Manufacturing or the person with the equivalent position in the company. An attached letter explained the purpose of the survey, the structure of the questionnaire, and assurances of confidentiality. Questionnaires were returned by mail or fax to the country office. Follow-up letters, faxes, emails, and calls helped to achieve a response rate close to 20%. Local researchers entered data from questionnaires into spreadsheets. The database was consolidated at Politecnico di Milano, Italy and released to the network. Further information about the survey administration can be found in Voss and Blackmon (1998) and Frohlich and Westbrook (2001). The database included 474 manufacturers, 285 of which had usable responses for the purposes of this study (Table 1). ANOVA tests found no significant differences between this subset and the total sample regarding size, production process, or process layout (Table 2). Thus, usable responses might be viewed as a random sample of the complete population of respondents; using the complete response subset did not bias the conclusions from statistical analyses to follow.

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947 Table 1 Sample composition (n = 285) n

%

Country of origin Argentina Australia Belgium China Denmark Germany Hungary Ireland Italy Norway Spain Sweden The Netherlands United Kingdom

10 28 14 10 26 23 42 16 39 15 12 13 8 29

3.5 9.8 4.9 3.5 9.1 8.1 14.7 5.6 13.7 5.3 4.2 4.6 2.8 10.2

ISIC code 381 Metal products 382 Machinery 383 Electrical equipment 384 Transportation equipment 385 Scientific equipment

83 93 62 22 25

29.1 32.6 21.8 7.7 8.8

3.2. Measures 3.2.1. Dependent variables The study assessed performance improvements in four flexibility dimensions, namely volume flexibility (VOLFLEX), mix flexibility (MIXFLEX), time to market (TTM), and product customization ability (CUSTOM). The four variables represented the main ‘aggregated’ flexibility dimensions in frameworks by Slack (1987), Bartezzaghi and Turco (1989), Suarez et al. (1991), and Chen et al. (1992), i.e. volume flexibility, mix flexibility, and with TTM and CUSTOM representing the ‘design’ and ‘modification’ aspects of product flexibility. Respondents were asked to, ‘‘Please Table 2 Differences between used and unused responses Variables

IMSS sample Study sample p-Value

Size (number of employees) 633.93 (472) a 603.63 (285) .755 Production process (% of value added) Fabrication 55.69 (455) Assembly 44.31 (455)

53.95 (280) 46.05 (280)

.483 .483

Process layout (% of total volume) Job shop 37.26 (452) Cellular manufacturing 30.31 (452) Dedicated lines 32.43 (452)

37.67 (285) 30.22 (285) 32.11 (285)

.886 .970 .908

Sample sizes are in parentheses; cases with missing data were excluded analysis-by-analysis. a Excluding two outliers (z > 14).

937

indicate the amount of change of the following performance dimensions over the last three years’’. Responses were given in a five-point scale with endpoints ‘‘strongly deteriorated’’ (1) and ‘‘strongly improved’’ (5). The questionnaire transcript is in Appendix A. 3.2.2. Independent variables The survey assessed the level of implementation of several action programs corresponding to initiatives that Collins and Schmenner (1993) associated with simplicity and discipline. The survey question, scales, and constructs are transcribed in Appendix A. Respondents used a five-point scale with endpoints ‘‘none’’ (1) and ‘‘high’’ (5) to indicate the level of implementation of 11 different programs over the previous 3 years. The initial scales of SIMPLICITY and DISCIPLINE were defined by identifying action programs with strong relationship to the simplicity and discipline initiatives identified by Collins and Schmenner (1993). SIMPLICITY incorporated six items. Information and communication technologies (IT/ERP) and e-business configurations (E-commerce/E-business) concerned improved information exchange and visibility of stocks and material flows in the supply chain. Non-core outsourcing, Process focus/re-layout, and Pull production focused on lean production aspects including inventory reduction, pull system signals, greater visibility of the factory floor, product or part families, re-layout of work areas, and less wasted motion. Improved new product development (NPD speed) focused on the use of techniques toward product modularization, platform design, and standard footprints. DISCIPLINE incorporated five items. Improving equipment standards (Process equipment update) and productivity (Equipment productivity/TPM) related to preventive maintenance practices, and following standard set-up and operation methods. Quality improvement and control (Quality programs) related to quality circles, operator checking on quality, continuous improvement, and problem identification and solving. Workforce development and Environment and safety initiatives focused on worker training and cross-training, follow-up with personnel affected, and good housekeeping. The scales validity and reliability were tested before conducting further analyses. Factor analysis with principal components and Varimax rotation was used to test convergent and discriminant validity (Table 3). The analysis yielded two factors with eigenvalues greater than 1. Only one item (Workforce development) loaded on more than one factor and was dropped from the analysis. All the remaining items loaded on a single factor, which supports convergent and discriminant validity. Cronbach’s

938

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

alphas were .75 for SIMPLICITY and .70 for the revised scale of DISCIPLINE, which confirms their reliability. Table 4 presents a summary of the final items used for the dependent and independent variables of this study, and how they compare to items used by Collins et al. (1998). As discussed earlier, dependent variables here were assessed via perceptual measures of performance in traditional flexibility dimensions; moreover, simplicity and discipline included a mix of organization and technology-based programs, as opposed to the variables in Collins et al. (1998) that included mostly organization-based programs. Nevertheless, at least three items used in this study (Pull production, Process equipment update, and Quality programs) could be paired to specific items from the previous study (respectively, kanban, process capability, and quality vision). 3.3. Moderator variable Hypothesis 3 predicted that dedicated lines would moderate relationships between simplicity/discipline and flexibility. DEDLINE was measured by the level of manufacturing activity carried out in dedicated lines (as opposed to job shops and cellular layouts): ‘‘Please indicate to what extent your activity is organized in the following layout categories (indicate percentage of total volume): job shop_%; cellular layout_%; dedicated lines_%.’’ Therefore, high values in dedicated lines

Factor 1

Factor 2

SIMPLICITY (a = .75) Non-core outsourcing Pull production E-commerce/E-business Process focus/re-layout NPD speed IT/ERP

.702 .668 .644 .631 .555 .523

.009 .253 .085 .332 .329 .268

DISCIPLINE (a = .70)a Process equipment update Environment and safety Equipment productivity/TPM Quality programs Workforce development

.036 .120 .438 .364 .525

.708 .696 .632 .617 .533

Rotation sums of squared loadings Eigenvalue Percent of variance explained Cumulative variance explained

2.95 26.80 26.80

Present study Flexibility Mix flexibility Product customization ability Time to market Volume flexibility

Simplicity Pull productiona E-commerce/E-business IT/ERP Non-core outsourcing NPD speed Process focus/re-layout

Collins et al. (1998) Customer delivery time Inventory management New product introduction Product cycle time Total cycle time Warehousing ! Kanban Problem solving Supplier lead time Supplier relations

Discipline Process equipment updatea ! Process capability Quality programsa ! Quality vision Environment and safety Business process documentation Equipment productivity/TPM Defects Manufacturing strategy Performance measurement Scrap and rework Training a

Paired items.

corresponded to high volume/line flow processes, and low values corresponded to low volume/intermittent processes. 3.4. Control variables

Table 3 Factor analysis-independent variables (n = 285) Variables

Table 4 Research items used in the two studies

2.41 21.93 48.73

Principal components analysis with Varimax rotation. Factor loadings greater than .5 are in boldface. a Excluding Workforce development.

The analysis incorporated three control variables that likely share variance with flexibility performance and the implementation of improvement programs. Organizational size can affect a unit’s performance since large firms tend to have more resources to invest in performance improvements (Tsai, 2001). SIZE was measured by the number of employees in the business unit: ‘‘At the end of the last fiscal year, in your business unit you had ___ employees in total’’. Market growth can have a negative impact in flexibility performance, as higher growth rate will lead the firm to invest less in flexibility (Tannous, 1996). GROWTH was assessed by the question, ‘‘How would you describe the development of the total market of that product line that you serve? (1) declining rapidly, (2) declining, (3) stable, (4) growing, (5) growing rapidly’’. Finally, capital investment in process equipment is often expected to yield flexibility improvements (Ghemawat and del Sol, 1998). EQUIPINV was measured through the question, ‘‘During the last three years, approxi-

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

mately what proportion of business unit revenues was spent on (average % of total revenues)?: ___% process equipment’’.

939

regression models. These tests were performed in step 3 of the regression models by entering the linear term of DEDLINE, the linear interaction of DEDLINE  SIMPLICITY or DEDLINE  DISCIPLINE, the quadratic term of DEDLINE, and the quadratic-bylinear term of interaction. For example, the final models having CUSTOM as dependent variable were defined as

4. Results Table 5 presents the means, standard deviations, and correlations between variables. Significant correlations between dependent variables indicate that sample companies often achieved simultaneous improvements in more than one flexibility type. The significant correlation between the two independent variables suggests that companies frequently implemented simplicity and discipline programs concurrently. There were no significant correlations between control variables. Hypotheses1 and 2 were tested with hierarchical regression analyses using the four flexibility performance variables of CUSTOM, VOLFLEX, MIXFLEX, and TTM as dependent variables. In step 1, SIZE, GROWTH, and EQUIPINV were entered as control variables; in step 2, either SIMPLICITY or DISCIPLINE was entered as independent variable. For example, the models having CUSTOM as dependent variable were defined as

CUSTOM ¼ b0 þ b1 SIZE þ b2 GROWTH þ b3 EQUIPINV þ b4 SIMPLICITY þ b6 DEDLINE þ b7 ðDEDLINE  SIMPLICITYÞ þ b9 DEDLINE2 þ b10 ðDEDLINE 2  SIMPLICITYÞ þ e CUSTOM ¼ b0 þ b1 SIZE þ b2 GROWTH þ b3 EQUIPINV þ b5 DISCIPLINE þ b6 DEDLINE þ b8 ðDEDLINE  DISCIPLINEÞ þ b9 DEDLINE2

CUSTOM ¼ b0 þ b1 SIZE þ b2 GROWTH

þ b11 ðDEDLINE 2  DISCIPLINEÞ þ e

þ b3 EQUIPINV þ b4 SIMPLICITY þ e CUSTOM ¼ b0 þ b1 SIZE þ b2 GROWTH

Tables 6 and 7 present the results of the regression analyses. Analysis of variance and change statistics between regression steps were used to estimate the significance of regression coefficients. All variables were normalized to minimize multi-colinearity problems and enable the comparison of parameter estimates across the models (Aiken and West, 1991). Individual tests were carried out for each flexibility type. All of the variance

þ b3 EQUIPINV þ b5 DISCIPLINE þ e Hypothesis 3 on the moderating effect was assessed through analytical procedures described in Cohen and Cohen (1983), Baron and Kenny (1986), and Aiken and West (1991). Moderation was tested by introducing linear and quadratic-by-linear interactions in the Table 5 Means, standard deviations, and bivariate correlations (n = 285) Variables

Mean

S.D.

1

2

3

4

5

6

7

8

9

1. CUSTOM 2. VOLFLEX 3. MIXFLEX 4. TTM 5. SIMPLICITY 6. DISCIPLINE 7. DEDLINE 8. SIZE 9. GROWTH 10. EQUIPINV

3.53 3.78 3.59 3.47 2.69 3.05 32.11 603.63 3.33 10.28

.75 .75 .74 .72 .81 .81 35.31 1234.57 .87 14.73

.11 .21 ** .16 ** .12 .15 * .02 .10 .02 .00

.51 ** .18 ** .18 ** .24 ** .03 .05 .04 .03

.29 ** .28 ** .26 ** .00 .06 .03 .01

.17 ** .12 * .03 .13 * .06 .07

.59 ** .03 .18 ** .05 .07

.05 .23 ** .01 .13 *

.16 ** .10 .05

.11 .03

.02

* **

p < .05 (two-tailed). p < .01 (two-tailed).

940

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

Table 6 Results of hierarchical regression analyses for flexibility performance on SIMPLICITY and DEDLINE (n = 285) Variablea

CUSTOM

VOLFLEX

Step 1 Step 2 Step 3

Step 1 Step 2

Intercept .000 SIZE .097 GROWTH .010 EQUIPINV .003 SIMPLICITY DEDLINE DEDLINE  SIMPLICITY DEDLINE SQUARED DEDLINE SQUARED  SIMPLICITY R2 .010 F-change .919 a * ** ***

.000 .079 .007 .010 .101

.120 .103 .000 .022 .264** .094 .051 .113 .150

.000 .050 .039 .029

.020 2.809

.046 1.907

.005 .497

MIXFLEX Step 3

Step 1 Step 2

.000 .006 .000 .019 .041 .059 .033 .017 .022 .018 .013 .008 .174** .398*** .043 .252** .010 .228** .034 8.459**

.071 2.682*

TTM

.004 .399

Step 3

Step 1

Step 2

Step 3

.000 .078 .000 .000 .012 .104 .116 .009 .017 .131* .013 .012 .048 .043 .041 .026 .023 .076 .086 .089 .283*** .288** .151* .237* .051 .024 .008 .043 .078 .017 .010 .086 .081 23.473***

.085 .274

.026 2.515

.048 6.431*

.054 .412

Standardized variables. p < .05. p < .01. p < .001.

inflation factors were below the cutoff of 10, suggesting that multi-colinearity was not a problem.

suggest that increasing investments in simplicity programs yielded flexibility benefits to manufacturing companies.

4.1. Simplicity 4.2. Discipline Hypothesis 1 predicted that simplicity programs would relate positively to flexibility. The results supported Hypothesis 1. As indicated in Table 6, the SIMPLICITY term entered in step 2 of the models appeared to relate positive and significantly to three of the four flexibility variables, namely VOLFLEX ( p < .01), MIXFLEX ( p < .001), and TTM ( p < .05). The results

Hypothesis 2 predicted that discipline programs would relate positively to flexibility. As indicated in Table 7, the regression analyses supported Hypothesis 2. Entering the DISCIPLINE term in step 2 of the flexibility models yielded positive and significant correlations with CUSTOM ( p < .05), VOLFLEX ( p < .001), and

Table 7 Results of hierarchical regression analyses for flexibility performance on DISCIPLINE and DEDLINE (n = 285) Variablea

CUSTOM Step 1 Step 2

Intercept .000 .000 SIZE .097 .067 GROWTH .010 .012 EQUIPINV .003 .020 DISCIPLINE .135* DEDLINE DEDLINE  DISCIPLINE DEDLINE SQUARED DEDLINE SQUARED  DISCIPLINE R2 .010 .027 F-change .919 4.931* a * ** ***

Standardized variables. p < .05. p < .01. p < .001.

VOLFLEX

MIXFLEX

Step 3

Step 1 Step 2

Step 3

.145 .080 .007 .039 .363*** .097 .196* .131 .218**

.000 .050 .039 .029

.000 .017 .000 .003 .026 .059 .043 .014 .022 .000 .012 .008 .236*** .486*** .044 .236** .000 .252***

.063 2.632*

.005 .497

.057 15.405***

.096 2.956*

Step 1 Step 2

TTM Step 3

Step 1

Step 2 Step 3

.000 .069 .000 .000 .006 .107 .117 .001 .016 .131* .026 .018 .048 .050 .039 .041 .044 .076 .089 .096 .267*** .360*** .106 .237* .045 .028 .041 .121 .074 .015 .095 .133

.004 .071 .399 20.041***

.081 .807

.026 2.515

.037 3.047

.048 .789

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

MIXFLEX ( p < .001). The results indicate that discipline programs were positively associated to flexibility in the sample. 4.3. Dedicated lines Hypothesis 3 predicted that the use of dedicated lines would moderate relationships between flexibility, and simplicity and discipline. As explained earlier, linear and quadratic-by-linear terms of interaction were introduced in step 3 of the models to test this hypothesis. The linear interaction term was significantly positive and the quadratic-by-linear interaction term was significantly negative in three of the six models that had shown significant direct relationships between independent and dependent variables, namely the SIMPLICITY–VOLFLEX model, the DISCIPLINE–CUSTOM model, and the DISCIPLINE– VOLFLEX model. The positive linear interaction term suggests that simplicity or discipline had a higher flexibility impact in high volume processes than in low volume processes. The negative quadratic-by-linear interaction term suggests that such impact was also higher in the middle range processes than in the extremes. Thus, in the case of SIMPLICITY–VOLFLEX, DISCIPLINE–CUSTOM, and DISCIPLINE– VOLFLEX, companies to benefit most from investments in simplicity/discipline appeared to be, first, those with medium-to-high volume processes, and second, those with high and medium-to-low volume processes; low-end process companies appeared to achieve the lowest flexibility benefits. The results provided partial support to Hypothesis 3, as the use of dedicated lines appeared to moderate some, but not all of the significant relationships between flexibility, and simplicity and discipline. To facilitate the interpretation of these terms and their coefficients, the relationships between simplicity, discipline and the four dependent variables of flexibility were graphed for different values of the independent and moderating variables, pursuant to the rationale in Baron and Kenny (1986) and Aiken and West (1991). Using the equations generated in step 3 of the models, several combinations of SIMPLICITY and DISCIPLINE (varying from 1 to +1 standard deviations from the mean), and low (plotted at .9 standard deviations from the mean) and high (plotted at +.9 standard deviations from the mean) DEDLINE were used to graph the relationships between the independent variables and performance at high and low volume processes (Fig. 2). The plots of volume flexibility were compatible with Hypothesis 3, as both

941

SIMPLICITY and DISCIPLINE appeared to have a positive effect only in the case of high volume processes. On the other hand, the plots of mix flexibility run counter to Hypothesis 3, as SIMPLICITY and DISCIPLINE appeared to yield positive performance improvements in both low volume and high volume processes. Implications from the other graphs were less straightforward. The moderating effect of DEDLINE on the DISCIPLINE–CUSTOM relationship could be easily identified in the product customization chart; however, the relationship between SIMPLICITY and CUSTOM appeared not significant and not moderated by DEDLINE, as both LV-S and HV-S had a small and positive slope. Finally, the time to market chart would initially suggest a limited moderating effect of DEDLINE on the relationship between DISCIPLINE and TTM. However, since neither the direct effect of DISCIPLINE on TTM nor the interaction effects of DEDLINE could be validated at the .05 level, the TTM models could not provide significant support to Hypothesis 3. In summary, the regression analyses provided evidence to partially support Hypothesis 3, as the use of dedicated lines in high volume processes appeared to moderate some, but not all, of the significant relationships between the independent and dependent variables in study. 5. Discussion The results indicated that building simplicity and discipline in manufacturing related positively to flexibility improvements, and that some relationships were more positive in high volume processes than in low volume processes. Moreover, the factor analysis validated the development of discipline/simplicity scales that integrated technology and organizationfocused initiatives. Thus, this study provides not only further empirical support to the rigid flexibility model, but also new knowledge on the model’s validity across different process types and using a socio-technical set of simplicity and discipline programs. 5.1. Flexibility performance This study reinforces the empirical validation in Collins et al. (1998) of positive relationships between simplicity/discipline and flexibility. However, while the previous work used measures such as inventory turnover, storage space, and production lead time to assess the plant flexibility, this study focused on perceptual measures of performance relating directly to the flexibility dimensions defined by Slack (1987) and Suarez et al. (1991), among others.

942

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

Fig. 2. Regression plots for flexibility on simplicity at low volume (LV-S) and high volume (HV-S) processes, and for flexibility on discipline at low volume (LV-D) and high volume (HV-D) processes. Step 3 coefficients are used. All variables are standardized. Control variables were set to zero. Moderation variable was set to .9 in low volume and +.9 in high volume cases.

Both simplicity and discipline appeared to relate positively to three of the four flexibility variables. Simplicity appeared to yield positive improvements in volume flexibility, mix flexibility, and time to market. Even though the study did not search for relationships between dependent variables and the individual items in the independent scales, one could conjecture that volume flexibility improvements might be particularly associated to the use of pull production, and more explicitly to the outsourcing that Jack and Raturi (2002) named as one of the best long-term sources of volume flexibility. Mix flexibility would have probably benefited from pull production and process focus approaches, while time to market could be linked to improved new product development and the use of information technologies and e-business configurations to promote concurrent engineering and knowledge about customer requirements and market trends. Discipline was positively associated to improvements in product customization, volume flexibility, and

mix flexibility. It seems that all of the four items in the discipline scale would enable the technical and human elements of the organization to respond to changes in demand in a reliable and timely fashion. Process equipment updates and Total Productive Maintenance might bring in process technologies that were more robust and capable of coping with innovative, smallquantity, and technically challenging components. Quality programs and Environment and safety initiatives would similarly enable workers to guarantee reliability and efficiency in a context of customized production or a broader product mix. 5.2. Interaction with dedicated lines The study tested the hypothesis of a moderating role of dedicated line layouts in relationships between simplicity/discipline and flexibility. As illustrated in Fig. 2, the analysis provided evidence to a moderating role of dedicated lines in three of the six significant

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

relationships, i.e. between simplicity and volume flexibility, discipline and product customization, and discipline and volume flexibility. In all of those cases, the independent variable appeared to have a more positive effect on flexibility in high volume processes than in low volume processes. Adding the results for the quadratic-by-linear interactions into this equation further suggested that this impact would be at its peak at middle-to-high volume processes, i.e. lower variety/ higher volume manufacturing cells and higher variety/ lower volume assembly or production lines. Therefore, it would appear that in the three cases with significant moderation, assembly/production line processes would have the most to benefit from simplicity and discipline, with continuous flow and batch processes achieving moderate benefits, and job shop and projecttype processes realizing the lowest results. The seemingly limited impact of simplicity and discipline programs in the flexibility of low volume processes can be explained by studies, e.g. Fisher (1997), Huang et al. (2002) that suggest that companies operating in market-responsive supply chains still need to deploy significant work-in-process and capacity buffers to cope with changes in customer demand. Moreover, flexibility in low volume/high variety processes may often depend more on sheer manufacturing skill and engineering expertise than on extra automation or process streamlining. As stressed by Kotha (1996, p. 851), in complex manufacturing situations, ‘‘Although the robots and semi-automated machines perform the rough and more touring tasks, it is craftsmen who complete the finest details’’. 5.3. The sociotechnical approach revisited Another innovation in the study has been the development of scales for simplicity and discipline that combined organization and technology-based initiatives. As discussed earlier, the simplicity scale involved technology programs such as IT/ERP, and organization methods such as process focus/re-layout. The discipline scale similarly involved technology programs such as Process equipment update, and organization methods such as Environment and safety. Given the significant levels of relationship between the new scales and flexibility, it seems appropriate to suggest that the simultaneous development of technical and organizational improvement practices can be at least as effective, or perhaps even more, than an exclusive focus on either technical or organizational improvements. This proposition is consistent with principles associated to the sociotechnical approach.

943

The sociotechnical approach emerged in the 1950s from Trist and Bamforth’s (1951) innovative study of the mining industry, and the work of the Tavistock Institute (Thompson and McHugh, 1990). Its basic premise is that operations require joint optimization of the technical system, with its technologies and equipment, and the social system, with workers, their skills and relationships (Moldaschl and Weber, 1998; Cronshaw and Alfieri, 2003). Common sociotechnical systems today include autonomous work groups, cell assembly manufacturing, and just-in-time work design (Lowenberg and Conrad, 1998; Hummels and de Leede, 2000). The approach’s major implication to operations improvements is that changes in the technology or equipment should be complemented by changes in the organizational structure (Cronshaw and Alfieri, 2003), as workers must have access to the skills and human resource support systems to enable performance with the new technical system. Therefore, the sociotechnical approach appears consistent with the view that the development of simplicity and discipline should simultaneously embrace technical and work-oriented improvement practices. This was pointed out by Gerwin (1988), who proposed that without compatibility between the innovation’s technical complexity and the organization’s augmented infrastructure, the innovation’s impact on flexibility would suffer. As he suggested in a later study (Gerwin, 1993), a variety of technology-based and organization-based methods are available to meet the needs for flexibility, but the real challenge for managers is how to properly balance these methods to deliver their potential flexibility benefits. Thus, the results from this study appear to confirm the view that flexibility may further benefit from balancing the technical and social aspects of manufacturing innovations. Given the importance of these findings, and even the controversy surrounding the sociotechnical approach in the recent literature (e.g. Moldaschl and Weber, 1998; Hummels and de Leede, 2000), there appears to be an opportunity for further research to assess the benefits and limitations of this approach in the context of operations improvements. 6. Conclusion The study tested the rigid flexibility model proposed by Collins and Schmenner (1993). Regression analysis was used to assess relationships between simplicity/ discipline and flexibility performance. The analysis indicated that simplicity and discipline were positively associated to flexibility improvements, and that the use of dedicated lines moderated some of those relationships. The results provide further empirical validation to

944

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

the rigid flexibility model. Moreover, they indicate that simplicity and discipline may yield greater benefits in high volume processes than in low volume processes. This appears to be the second study ever to test the rigid flexibility model. Compared to the pioneer research by Collins et al. (1998), this study introduced some innovations to explore further the model’s validity and generalizability in manufacturing. The dependent variables incorporated performance measures with direct representation to the main flexibility dimensions developed in the literature. Data included manufacturers from 14 countries. Finally, simplicity and discipline scales incorporated items focused not only on organizational improvements but also on the information and process technologies that today appear critical to flexibility. At least three major limitations may be associated to this study. First, while the simplicity and discipline constructs involved a large variety of action programs, the flexibility construct was limited to ‘aggregated’ flexibility types; however, studies such as Browne et al. (1984) stressed the role of ‘individual resource’ dimensions such as machine and labor flexibility in manufacturing. The study approach can be justified by the rationale in Slack (1987) and Sethi and Sethi (1990) that low-level flexibility dimensions are embedded in high-level dimensions. Second, data were obtained from an existing database rather than a study-focused survey. Using the IMSS database can be justified by the gains in sample reliability and scale it promotes. Third, data was limited to manufacturers of fabricated metal products, machinery, and equipment. Focus on a single industry is known to maximize validity but limit the generalizability of research findings (Ketchen et al., 1993). This study has practical implications deriving from the original model and the research findings. Studies such as Swamidass and Newell (1987), Fiegenbaum and Karnani (1991), and Narasimhan and Das (1999) provided evidence that flexibility improvements often translate into operational and financial performance. What this study suggests is that such flexibility improvements, especially in high volume processes, can be anchored in simplicity and discipline in operations. Alternative approaches such as building inventory cushions and buffer capacity may bring in more flexibility for the short term but, except for companies producing the highest levels of variety, they may also increase cost and complexity to a point that will cripple the organization’s performance in the long term. The study also suggests that simplicity and discipline may incorporate simultaneous improve-

ments in technology and work methods, which reinforces the merits of using a sociotechnical approach to operations investments. Hence, this study has some implications for further research, the first of which concerns the opportunity to explore the merits of a sociotechnical, joint optimization approach to operations design and improvement. As discussed in the previous section [and reinforced in studies such as Moldaschl and Weber, 1998; Hummels and de Leede, 2000], the sociotechnical approach appears to provide as many answers as challenges to business management. Despite its clear implications to research and practice, operations management studies still appear to lag other disciplines in this issue. Second, results indicated that the rigid flexibility model might support improvements in mix flexibility and time to market in both high volume and low volume processes. However, only high volume processes appeared to benefit in terms of product customization and volume flexibility. Therefore, further research is needed to assess the drivers of product customization and volume flexibility in low volume processes. Appendix A. Scales and constructs A.1. Flexibility (y1, y2, y3, y4) D2. Please indicate the amount of change of the following performance dimensions over the last 3 years:

Product customization ability Volume flexibility Mix flexibility Time to market

Strongly deteriorated

No change

Strongly improved

1

2

3

4

5

1 1 1

2 2 2

3 3 3

4 4 4

5 5 5

A.2. Simplicity (x1). Cronbach’s alpha = .75 C5. This question explores the action programs1 to which your company is now devoting high resource and innovation effort and on which is concentrated the management focus and commitment. On the far left side, indicate whether the program has been undertaken within the last 3 years.

1 By action program is meant a major project aimed at producing considerable changes in the company’s management practices and organization.

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

945

Degree of use last 3 years None

High

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

Implementing Information and Communication Technologies and/or Enterprise Resource Planning software (IT/ERP) Reorganizing your company towards e-commerce and/or e-business Configurations (E-commerce/E-business) Concentrating on your core activities and outsourcing support processes and activities (e.g. IS management, maintenance, material handling, etc.) (Non-core outsourcing) Restructuring your manufacturing processes and layout to obtain process focus and streamlining (e.g. reorganize plant-within -a-plant; cellular layout, etc.) (Process focus/re-layout) Undertaking actions to implement pull production (e.g. reducing batches, setup time, using kanban systems, etc.) (Pull production) Implementing actions to improve or speed-up you process of new product development through e.g. platform design, products modularization, components standardization, concurrent engineering, Quality Function Deployment, etc. (NPD speed)

A.3. Discipline (x2). Cronbach’ alpha = .70 Same question as above. Degree of use last 3 years None

High

1 1

2 2

3 3

4 4

5 5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

a

Updating your process equipment to industry standard or better (Process equipment update) Undertaking programs for quality improvement and control (e.g. TQM programs, 6s projects, quality circles, etc.) (Quality programs) Undertaking programs for the improvement of your equipment productivity (e.g. Total Productive Maintenance programs) (Equipment productivity/TPM) Implementing actions to increase the level of delegation and knowledge of your workforce (e.g. empowerment, training, improvement or autonomous teams, etc.) (Workforce development)a Putting efforts and commitment on the improvement of your company’s environmental compatibility and workplace safety and health (Environment and safety)

Dropped from final scale.

A.4. Dedicated lines (moderator)

a. ___ Employees in total.

PT2. Please indicate to what extent your activity is organized in the following layout categories: (indicate percentage of total volume)

A.6. Market growth (control) A8. How would you describe the development of the total market of that product line that you serve?

Process layout Declining rapidly Declining Stable Growing Growing rapidly Job shop Cellular layout* Dedicated lines

___% ___% ___% 100%

*

Note: A ‘‘cell’’ is a grouping of equipment dedicated to support the production of families of parts sharing similar process operations.

A.5. Size (control) O1. At the end of the last fiscal year, in your business unit you had:

1

2

3

4

5

A.7. Equipment investment (control) D3. During the last 3 years, approximately what proportion of business unit revenues was spent on (average % of total revenues): ___% Process equipment.

946

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947

References Adler, P., 1985. Managing flexibility: a selective review of the challenges of managing the new production technologies’ potential for flexibility. Report to the Organizational for Economic Cooperation and Development, Stanford University, 36 pp. ˚ hlstro¨m, P., Westbrook, R., 1999. Implications of mass customizaA tion for operations management: an exploratory survey. International Journal of Operations and Production Management 19 (3), 262–274. Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and Interpreting Interactions. Sage, Newbury Park, CA. Barnes-Schuster, D., Bassok, Y., Anupindi, R., 2002. Coordination and flexibility in supply contracts with options. Manufacturing and Service Operations Management 4 (3), 171–207. 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), 1173–1182. Bartezzaghi, E., Turco, F., 1989. The impact of just-in-time on production system performance: an analytical framework. International Journal of Operations and Production Management 9 (8), 40–62. Bayus, B.L., Putsis Jr., W.P., 1999. Product proliferation: an empirical analysis of product line determinants and market outcomes. Marketing Science 18 (2), 137–153. Beach, R., Muhlemann, A.P., Price, D.H.R., Paterson, A., Sharp, J.A., 2000. Manufacturing operations and strategic flexibility: survey and cases. International Journal of Operations and Production Management 20 (1), 7–30. Bordoloi, S.K., Cooper, W.W., Matsuo, H., 1999. Flexibility, adaptability, and efficiency in manufacturing systems. Production and Operations Management 8 (2), 133–150. Boynton, A.C., Victor, B., Pine II, B.J., 1993. New competitive strategies: challenges to organizations and information technology. IBM Systems Journal 32 (1), 40–64. Browne, J., Dubois, D., Rathmill, K., Sethi, S.P., Stecke, K.E., 1984. Classification of flexible manufacturing systems. The FMS Magazine 2 (2), 114–117. Chen, I.J., Calantone, R.J., Chung, C.H., 1992. The marketing-manufacturing interface and manufacturing flexibility. Omega 20 (4), 431–443. Cohen, J., Cohen, P., 1983. Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences, second ed. Erlbaum, Hillsdale, NJ. Collins, R.S., Cordon, C., Julien, D., 1998. An empirical test of the rigid flexibility model. Journal of Operations Management 16 (2/ 3), 133–146. Collins, R.S., Schmenner, R.W., 1993. Achieving rigid flexibility: factory focus for the 1990s. European Management Journal 11 (4), 443–447. Cronshaw, S.F., Alfieri, A.J., 2003. The impact of sociotechnical task demands on use of worker discretion and functional skill. Human Relations 56 (9), 1107–1130. De Toni, A., Tonchia, S., 1998. Manufacturing flexibility: a literature review. International Journal of Production Research 36 (6), 1587– 1617. Fiegenbaum, A., Karnani, A., 1991. Output flexibility—a competitive advantage for small firms. Strategic Management Journal 12 (2), 101–114. Fisher, M.L., 1997. What is the right supply chain for your product? Harvard Business Review 75 (2), 105–116.

Flynn, B.B., Flynn, E.J., 1999. Information-processing alternatives for coping with manufacturing environment complexity. Decision Sciences 30 (4), 1021–1052. Frohlich, M.T., Westbrook, R., 2001. Arcs of integration: an international study of supply chain strategies. Journal of Operations Management 19 (2), 185–200. Gerwin, D., 1988. A theory of innovation processes for computeraided manufacturing technology. IEEE Transactions on Engineering Management 35 (2), 90–100. Gerwin, D., 1993. Manufacturing flexibility: a strategic perspective. Management Science 39 (4), 395–410. Ghemawat, P., del Sol, P., 1998. Commitment versus flexibility? California Management Review 40 (4), 26–42. Gupta, Y.P., Goyal, S., 1989. Flexibility of the manufacturing system: concepts and measurement. European Journal of Operations Research 43 (2), 119–135. Huang, S.H., Uppal, M., Shi, J., 2002. A product driven approach to manufacturing supply chain selection. Supply Chain Management 7 (4), 189–199. Hummels, H., de Leede, J., 2000. Teamwork and morality: comparing lean production and sociotechnology. Journal of Business Ethics 26 (1), 75–88. Jack, E.P., Raturi, A., 2002. Sources of volume flexibility and their impact on performance. Journal of Operations Management 20 (5), 519–548. Karuppan, C.M., Ganster, D.C., 2004. The labor-machine dyad and its influence on mix flexibility. Journal of Operations Management 22 (6), 533–556. Kekre, S., Srinivasan, K., 1990. Broader product line: a necessity to achieve success? Management Science 36 (10), 1216–1231. Ketchen Jr., D.J., Thomas, J.B., Snow, C.C., 1993. Organizational configurations and performance: a comparison of theoretical approaches. Academy of Management Journal 36 (6), 1278– 1313. Kotha, S., 1996. Mass-customization: a strategy for knowledge creation and organizational learning. International Journal of Technology Management 11 (7/8), 846–858. Lowenberg, G., Conrad, K.A., 1998. Current Perspectives in Industrial/Organizational Psychology. Allyn & Bacon, Boston, MA. Mendelson, H., Pillai, R.R., 1999. Industry clockspeed: measurement and operational implications. Manufacturing and Service Operations Management 1 (1), 1–20. Moldaschl, M., Weber, W.G., 1998. The ‘three waves’ of industrial group work: historical reflections on current research on group work. Human Relations 51 (3), 347–388. Narasimhan, R., Das, A., 1999. An empirical investigation of the contribution of strategic sourcing to manufacturing flexibilities and performance. Decision Sciences 30 (3), 683– 718. Noble, M.A., 1995. Manufacturing strategy: testing the cumulative model in a multiple country context. Decision Sciences 26 (5), 693–721. Pagell, M., Krause, D.R., 1999. A multiple-method study of environmental uncertainty and manufacturing flexibility. Journal of Operations Management 17 (3), 307–325. Pagell, M., Krause, D.R., 2004. Re-exploring the relationship between flexibility and the external environment. Journal of Operations Management 21 (6), 629–649. Sethi, A.K., Sethi, S.P., 1990. Flexibility in manufacturing: a survey. International Journal of Flexible Manufacturing Systems 2 (4), 289–328.

G.J.C. da Silveira / Journal of Operations Management 24 (2006) 932–947 Slack, N., 1983. Flexibility as a manufacturing objective. International Journal of Operations and Production Management 3 (3), 4–13. Slack, N., 1987. The flexibility of manufacturing systems. International Journal of Operations and Production Management 7 (4), 35–45. Small, M.H., Chen, I.J., 1997. Organizational development and time-based flexibility: an empirical analysis of AMT adoptions. International Journal of Production Research 35 (11), 3005– 3021. Suarez, F.F., Cusumano, M.A., Fine, C.H., 1991. Flexibility and performance: a literature critique and strategic framework, Working paper #3298-91-BPS, Sloan School of Management, Cambridge, MA, 47 pp. Suarez, F.F., Cusumano, M.A., Fine, C.H., 1996. An empirical study of manufacturing flexibility in printed circuit board assembly. Operations Research 44 (1), 223–240. Swamidass, P.M., Newell, W.T., 1987. Manufacturing strategy, environmental uncertainty and performance: a path analytic model. Management Science 33 (4), 509–524. Tannous, G.F., 1996. Capital budgeting for volume flexible equipment. Decision Sciences 27 (2), 157–184.

947

Thilander, M., 1992. Flexible production in the chemical industry—a question of competence. International Journal of Operations and Production Management 12 (7/8), 147–167. Thompson, P., McHugh, D., 1990. Work Organisations—A Critical Introduction. MacMillan, Basingstoke, Hampshire. Trist, E.L., Bamforth, K.W., 1951. Some social and psychological consequences of the longwall method of coal-getting. Human Relations 4 (1), 3–38. Tsai, W., 2001. Knowledge transfer in intraorganizational networks: effects of network position and absorptive capacity on business unit innovation and performance. Academy of Management Journal 44 (5), 996–1004. Upton, D.M., 1994. The management of manufacturing flexibility. California Management Review 36 (2), 72–89. Van Dijk, M., 1995. Flexible specialisation, the new competition and industrial districts. Small Business Economics 7 (1), 15–27. Voss, C., Blackmon, K., 1998. Differences in manufacturing strategy decisions between Japanese and Western manufacturing plants: the role of strategic time orientation. Journal of Operations Management 16 (2/3), 147–158.