Manufacturing strategy in context: environment, competitive strategy and manufacturing strategy

Ž .Journal of Operations Management 18 2000 123–138www.elsevier.comrlocaterdsw

Manufacturing strategy in context: environment, competitivestrategy and manufacturing strategy

Peter T. Ward a,), Rebecca Duray b,1

a Max M. Fisher College of Business, Ohio State UniÕersity, Columbus, OH 43210, USAb College of Business and Administration, UniÕersity of Colorado at Colorado Springs, Colorado Springs, CO 80933-7150, USA

Received 1 July 1997; accepted 1 March 1999

Abstract

Considering manufacturing strategy in its larger strategic context has been thematic in conceptual literature in operationsbut relatively neglected in empirical studies, thus leaving predominant conceptual models of manufacturing strategy largelyuntested. This research develops a conceptual model of manufacturing strategy from the literature and tests the model usingdata from a sample of manufacturers in three industries in the United States. This research contributes to manufacturingstrategy literature in four ways. First, it supports empirically a model of manufacturing strategy that is predominant in theconceptual literature. Second, it demonstrates that the strategic linkages in manufacturing businesses are clearer among goodperformers than poor performers. Third, this research suggests that competitive strategy acts as a mediator between anorganization’s environment and its manufacturing strategy. Fourth, the findings suggest that the relationship betweencompetitive strategy and performance is mediated by manufacturing strategy. These last two findings have importantimplications for approaching research in manufacturing strategy in the future. q 2000 Elsevier Science B.V. All rightsreserved.

Keywords: Operations strategy; Empirical research

1. Introduction

Research in operations management has beencharacterized in recent years by an increasing effortdevoted to the study of manufacturing strategy usingempirical methods. A review of the literature revealsthat much of this empirical research effort has fo-

) Corresponding author. Tel.: q1-614-292-5294; E-mail:[email protected]

1 Tel.: q1-719-262-3673; E-mail: [email protected].

cused on the internal consistency of manufacturingŽ .strategy e.g., priorities and programs and assessing

the performance consequences of such consistency.Surprisingly little empirical research has addressedthe alignment among manufacturing strategy, busi-ness-level competitive strategy, and the competitiveenvironment faced by the firm, although much of theconceptual literature in manufacturing strategy has

Žfocused on this issue of alignment Swink and Way,.1995 . Thus the predominant conceptual model of

manufacturing strategy that considers manufacturing

0272-6963r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0272-6963 99 00021-2

( )P.T. Ward, R. DurayrJournal of Operations Management 18 2000 123–138124

in the larger strategic context of the firm has re-mained largely unsubstantiated because it has notbeen adequately tested.

We approach this relatively neglected area bytesting empirically an accepted conceptual model ofmanufacturing strategy in the context of a sample offirms’ competitive strategies and environments. Inessence, we address three issues. First, we askwhether data collected from a sample of manufactur-ers are consistent with the model supported by muchof the conceptual literature. We describe that concep-tual model in Section 2. Second, we address whetheror not manufacturing strategy appears to matter inthe larger context of the firm’s environment andcompetitive strategy. In other words, we test whetherthere is a relationship between manufacturing strat-egy and business performance when the effects ofenvironment and business-level competitive strategyare also considered. Third, we address the form ofthe relationship between competitive environmentand manufacturing strategy. Specifically, we analyzethe extent to which competitive strategy mediates theeffects of environmental dynamism on manufactur-ing strategy.

Environmental dynamism refers to the degree ofturbulence in products, technologies, and demand for

Žproducts in a market Miller and Friesen, 1983; Dess.and Davis, 1984 . By competitive strategy we refer

to the broad dimensions that a business uses as abasis of advantage, e.g., price vs. differentiationŽ .Porter, 1980 . Manufacturing strategy may bethought of as the manufacturing-oriented dimensions

Ž .that win orders Hill, 1994 . Although the possiblemediating effects of competitive strategy on the rela-tionship between environmental dynamism and man-ufacturing strategy have not been tested previously,the environment has long been identified as an im-portant contingency in conceptual and empiricalstudies of both competitive and manufacturing strat-

Žegy e.g., Skinner, 1969; Hofer, 1975; Van Dier-.donck and Miller, 1980 .

To address these issues, we employ data from asample 101 U.S. manufacturers across three indus-tries to estimate a path model using covariance struc-ture analysis. We contrast the behavior of high andlow performers by splitting the sample on the basisof business performance. We utilize self-reportedperformance measures to achieve the performance

split and to analyze performance effects, although weacknowledge the shortcomings inherent in such data.

2. Manufacturing strategy model

A review of the literature in the area reveals thatthe conceptual model presented in Fig. 1 is a repre-sentative view of manufacturing strategy in its con-text. The model suggests that environmental dy-namism affects both competitive strategy and manu-facturing strategy. Competitive strategy is cast in amediating relationship because it intervenes betweenenvironmental dynamism and manufacturing strategyŽ .Venkatraman, 1989 . The model also implies thatcompetitive strategy directly influences manufactur-ing strategy. Further, the model suggests that the

Fig. 1. Conceptual model of manufacturing strategy in its context.

( )P.T. Ward, R. DurayrJournal of Operations Management 18 2000 123–138 125

relationship of environment, competitive strategy,and manufacturing strategy is linked to performance.The model also implies that direct links exist be-tween strategies and performance. The numbering ofthe arrows on Fig. 1 refers to hypotheses developedbelow.

2.1. Conceptual basis

The model in Fig. 1 is familiar to students ofmanufacturing strategy. The model’s origin can be

Ž .traced to the seminal paper of Skinner 1969 onmanufacturing strategy that prescribed in detail thelinks among environment, competitive strategy, andmanufacturing strategy to achieve good business per-formance. In their reviews of the operations strategy

Ž .literature, Anderson et al. 1989 and Leong et al.Ž .1990 find broad support for the conceptual modelintroduced by Skinner, although relatively little em-pirical evidence. A contemporary review reveals that

Ž .this finding still holds. For example, Hill 1994Ž .incorporates environment especially markets , com-

petitive strategy and manufacturing strategy in hisconceptual model, but the model is not tested empiri-cally. A review of the literature also reveals noinstance where the connections among environment,competitive and manufacturing strategy and perfor-mance implications are considered simultaneouslyusing empirical evidence.

Although no empirical study incorporates all thedimensions represented in Fig. 1, pieces of the con-ceptual model have been tested. For example, thelinks among competitive strategy, manufacturingstrategy, and performance are addressed by Vickery

Ž .et al. 1993 , who find covariance between competi-tive strategy and production competence with busi-ness performance. Another example of empirical re-search on the connection between competitive andmanufacturing strategies is the numerical taxonomyof manufacturing strategy types of Miller and RothŽ .1994 , which are found to be related to competitivestrategies, in some instances. In their study of firms

Ž .in the textile industry, Williams et al. 1995 find arelationship between competitive strategy and manu-facturing strategy and also between manufacturing

Ž .strategy and performance. Gupta and Lonial 1998use a path model to test linkages among businessstrategy, manufacturing strategy, and organizational

performance. None of these studies address the ef-fects of environment on strategy choice nor are therelationships considered simultaneously.

The linkages among environmental dynamism,manufacturing strategy and performance are ex-

Ž .plored empirically by Swamidass and Newell 1987Ž .and Ward et al. 1995 . Both papers show that

environmental dynamism is positively related tomanufacturing flexibility. The latter paper also findspositive links between environmental dynamism andquality and delivery capabilities among high per-formers. Both of these studies use path models toestablish that environmental factors affect manufac-turing strategy and performance, but neither includescompetitive strategy in the model.

In contrast with manufacturing strategy research,many empirical studies in competitive strategy havefound relationships among environment, competitive

Ž .strategy, and performance. Keats and Hitt 1988 usea covariance structure model to describe the relation-ship among several environmental dimensions, com-

Ž .petitive strategy, and performance. Miller 1988supports earlier conceptual work on the types ofenvironmental and strategic configurations that lead

Ž .to good performance. Kim and Lim 1988 alsoprovide evidence for the model linking environment,competitive strategy, and performance. In general,environmental dynamism is shown to be positively

Ž .linked with differentiation as opposed to cost-basedcompetitive strategies.

2.2. Hypotheses

Our literature review reveals that despite concep-tual support for a model linking environment, com-petitive strategy, manufacturing strategy and perfor-mance, simultaneous empirical consideration of allof these aspects has been lacking. We address thisdeficiency in the literature by considering first thesequential relationships in Fig. 1; those paths directlylinking environmental dynamism with competitive

Ž .strategy path 1a , competitive strategy with manu-Ž .facturing strategy path 1b , and manufacturing strat-

Ž .egy with performance path 1c . Simultaneously test-ing the existence of these three paths implicitly teststhe essence of the conceptual model predominant inthe operations strategy literature and addresses thequestion of whether the model is supported by the


data. We also test whether significant links existbetween environment and manufacturing strategyŽ .path 2 and between competitive strategy and per-

Ž .formance path 3 . In addition, we test whetherbusinesses that are high performers are more apt toconform to the conceptual model than lower per-formers.

2.2.1. Fit of the conceptual modelBecause of the broad support in the literature, we

expect that the model will fit for firms that exhibitrelatively high business performance. We distinguishbetween relatively high and low performers becausethe literature suggests that low performers are lesslikely to adhere to the model. For example, SkinnerŽ .1969 argues that manufacturing’s task has to fit thecompetitive strategy or failure is almost inevitable.

Ž .Similarly, Hill 1994 prescribes a strategic processaimed at avoiding misalignments between marketinggoals and manufacturing capabilities because suchmisalignments are frequently causes of poor businessperformance. Therefore, we expect high performersto adhere to the model and the behavior of lowperformers to fit the model poorly.

More specifically, we expect to find statisticalsupport for a good fit for a path model directlylinking environment, competitive strategy, manufac-turing strategy among high performers. This model isdepicted in paths 1a, 1b and 1c of Fig. 1. Theabsence of significant linkages or good model fitamong good performers would indicate that the datado not support this conceptual model.

Hypothesis 1: High performers conform to the con-ceptual model to a greater extent than low perform-ers.

2.2.1.1. EnÕironmental dynamism and competitiÕestrategy. In addition to overall model fit, we test therelationships suggested by each of the paths shownin Fig. 1. In each instance, the literature providessome evidence for the existence of a relationship.For example, the literature is replete with empiricalevidence that environment influences strategy. InSection 2, we note several such studies that report asignificant relationship between environmental dy-namism and competitive strategy. In general these

studies show that more dynamic environments sup-port competitive strategies that are based on differen-tiation. We expect similar findings from our analysis.

Hypothesis 1a: Environmental dynamism influencescompetitive strategy choice.

2.2.1.2. CompetitiÕe strategy and manufacturingŽ .strategy. Although Swink and Way 1995 point out

that they are relatively few in number, extant studiesdo provide empirical evidence of the existence of thepredicted linkage between competitive strategy andmanufacturing strategy, i.e., that manufacturing strat-egy supports competitive strategy in high performingbusinesses. For example, this relationship is borne

Ž .out by Vickery et al. 1993 . It is important to notethat a major tenet in the development of manufactur-ing strategy has been that poor business performanceresults when manufacturing strategy is not linkedwith competitive strategy. We test the followinghypothesis, expecting to identify a systematic rela-tionship between competitive strategy and manufac-turing strategy.

Hypothesis 1b: Competitive strategy influences man-ufacturing strategy.

2.2.1.3. Manufacturing strategy and performance.The existence of a relationship between manufactur-ing strategy and business performance has long beensupported by the manufacturing strategy literature.

Ž .For example, Swamidass and Newell 1987 showedthat performance was positively related to a particu-lar manufacturing strategy, flexibility. A number ofstudies have shown that quality is linked with goodperformance. For example, Ferdows and DeMeyerŽ . Ž .1990 and Noble 1995 have argued that effectivemanufacturing strategies generally begin with qualityas a base. Several studies describing world classmanufacturers suggest that the best competitors com-pete on the basis of a variety of manufacturing

Žcapabilities e.g., Flynn et al., 1995b; Ward et al.,.1996; Collins et al., 1998 . We test the following

hypothesis expecting a positive relationship betweenemphasis on manufacturing capabilities and businessperformance among high performance manufactur-ers.


Hypothesis 1c: Manufacturing strategy influencesbusiness performance.

2.2.2. EnÕironment and manufacturing strategyThe literature also contains evidence of a direct

relationship between environmental factors, in partic-ular environmental dynamism, and manufacturingstrategy. We earlier cited Swamidass and NewellŽ . Ž .1987 and Ward et al. 1995 as examples. In bothinstances, the researchers also linked business perfor-mance to the relationship, suggesting that the highperforming firms choose manufacturing strategiesconsistent with their environments. Because of thisevidence, we also test for the existence of a directrelationship between environmental dynamism andmanufacturing strategy in the presence of competi-

Ž .tive strategy Fig. 1, path 2 . The absence of asignificant coefficient for such a path would implythat competitive strategy mediates the effects of en-vironmental dynamism on manufacturing strategywith no significant independent effect.

Hypothesis 2: Environmental dynamism has a directinfluence on manufacturing strategy.

2.2.3. CompetitiÕe strategy and performanceThe principle of equifinality suggests that a num-

ber of different, equally effective strategies can beŽused to achieve good performance Van de Ven and

.Drazin, 1985; Doty et al., 1993 . Strategy implemen-tation is the key link between competitive strategyand success measured by business performance. Anumber of authors have argued that functional strat-egy, in general, or manufacturing strategy, in particu-lar, describes such implementation by providing amore detailed picture of how a competitive strategy

Ž .is pursued e.g., Hatten et al., 1978; Miller, 1987 .We test the significance of the path from competitivestrategy to performance to test the importance of theintervention of manufacturing strategy in definingthat relationship between competitive strategy and

Ž .performance Fig. 1, path 3 . The absence of signifi-cance for such a link would imply that manufactur-ing strategy mediates the relationship between com-petitive strategy and performance rather than havingan independent effect.

Hypothesis 3: Competitive strategy has a direct in-fluence on business performance.

3. Data and methods

3.1. Data

Data for the study are from a survey of U.S.manufacturing firms. The sample consists of busi-nesses whose primary product is in one of threesectors: fabricated metal products; electrical devices;and electronic controls. The sample is restricted tobusinesses reporting 150 or more employees at thelocation. Each firm in the sample was contacted bytelephone to identify the top manager resident at that

Ž .location for example, general manager and verifyaddress and line of business. Thus identified, generalmanagers were called and asked to participate in thestudy. Those who agreed to participate were asked to

Ž .provide names and addresses for their: i plantŽ . Ž .manager, ii marketing manager, and iii manufac-

turing engineering manager. Distinct survey formswere then mailed to each of the four managers sothat two independent responses could be elicited foreach question from each responding firm Thesemethods yielded 101 usable responses, a responserate of 37%.

Analysis of the variables used in this researchindicates no significant pattern of responses by in-dustry and, thus, no identifiable industry effects. Acheck for non-respondent bias was carried out bycomparing non-respondents with respondents in thesame industry on the basis of reported sales volumeand number of employees. This check revealed nopattern of differences between the two groups. Intelephone interviews with refusing firms, non-re-spondents’ lack of time and reluctance to revealconfidential information were most commonly citedas reasons for not participating.

Each perceptual question is asked of two differentmanagers at the same firm to allow a check onwhether each informant’s perception is shared byothers in the firm. Correlations between the tworespondents for each of the scales are all significantat less than 0.01, indicating a high degree of inter-rater agreement and lending support to the validity ofthe scales.


3.2. Operational definitions

The constructs in the research are made opera-tional using multi-item scales intended to capture theunderlying theoretical dimensions. For each con-struct, we describe briefly both the antecedents orconceptual underpinnings and mechanics used inmeasurement. The values for each scale are repre-sented as standardized scores. Specific scales areshown in Appendix A.

3.2.1. EnÕironmental dynamismThe environmental dynamism items are adapted

Ž .from Miller and Friesen 1983 . Respondents areasked to indicate the rate of change, from slow torapid, at which products and services become obso-lete, the rate of innovation in productrservice and inprocess, and the rate of change in customers’ tastesand preferences. High numeric scores indicate higherrates of dynamism.

3.2.2. CompetitiÕe strategyŽ .Porter 1980 developed the idea that all competi-

tive strategies are variants of generic strategies in-volving a choice between differentiation and deliv-

Ž .ered cost price , with degree of focus, i.e., servingniche or broad markets, providing a second competi-tive dimension. Although many authors have sinceembellished or modified Porter’s generic strategies,they remain useful anchors for classifying strategiesŽ . Ž .e.g., Miller, 1988 . Dess and Davis 1984 providean instrument to make Porter’s generic strategic typesoperational which has been used frequently in strat-egy research and is validated by Kotha and Vadla-

Ž .mani 1995 . We adapt the instrument and findingsŽ .of Dess and Davis 1984 to measure the price and

differentiation aspects of competitive strategy.Specifically, for cost measures we consider the im-portance to the business unit of operating efficiency,competitive pricing, procurement of raw materials,minimizing outside financing, decreased number ofproduct features, and reducing product costs. Fordifferentiation, we use the importance to the businessunit of new product development, brand identifica-tion, innovation in marketing techniques and meth-ods, innovation in products and services, advertising,reputation within the industry, and forecasting mar-ket growth. For both cost and differentiation, highernumeric scores indicate greater emphasis.

3.2.3. Operations strategy and competitiÕe prioritiesA common theme in operations strategy research

has been describing manufacturers’ choices of em-phasis among key capabilities or competitive priori-ties. As noted above, the manufacturing strategyliterature suggest four competitive priorities: low

Žcost, quality, delivery, and flexibility Van Dier-donck and Miller, 1980; Hayes and Wheelwright,

.1984; Wheelwright, 1984 . More recently, Youndt etŽ .al. 1996 used these dimensions to operationalize

manufacturing strategy. It should be noted that otherpriorities could be included, notably innovationŽ .Hayes et al., 1988 .

Ž .3.2.3.1. Quality. Garvin 1987 points out that qual-ity is multidimensional and that each of its dimen-sions can be used strategically to gain competitiveadvantage. The quality scale that we use includesitems related to the important quality aspect of pro-

Žcess control and process management Flynn et al.,.1994 . Specifically,the scale measures organizational

emphasis on statistical process control, real timeprocess control systems, updating process equip-ment, and developing new processes for new and oldproducts.

3.2.3.2. Flexibility. The flexibility scale is intendedto capture the importance of reducing costs associ-ated with changing products or mix. Specifically, thescale measures the relative emphasis placed on lead-time reductions, set-up time reductions, the ability tochange priority of jobs on the shop floor, and theability to change machine assignments on the shop

Ž .floor Gerwin, 1993 .

3.2.3.3. DeliÕery. This measure includes emphasison customer service as indicated by either deliveryreliability or delivery speed.

3.2.3.4. Low cost. The instrument captures the com-petitive priority of low cost by measuring the empha-sis placed on reducing production costs, reducinginventory, increasing equipment utilization, and in-creasing capacity utilization.

3.2.4. PerformanceBusiness performance is operationalized as a

composite of two measures, market share and sales


Table 1Ž .Correlation matrix for full sample Cronbach’s coefficient alphas on the diagonal

Scale 1 2 3 4 5 6 7

EnÕironment1. Environmental dynamism 0.86

CompetitiÕe strategyUU2. Marketing differentiation 0.54 0.87

3. Price differentiation y0.09 y0.04 0.82

Manufacturing strategyU UU4. Low cost priority 0.08 0.24 0.29 0.67

U UUq5. Quality priority 0.21 0.20 0.02 0.44 0.72UU UU q6. Delivery priority 0.10 0.30 0.13 0.34 0.18 0.78

U UU UU UU UU7. Flexibility priority 0.21 0.31 0.06 0.46 0.43 0.30 0.73

USignificant at -0.05.

UUSignificant at -0.01.

qSignificant at -0.10.

growth, relating to the business unit’s perceivedŽ .position in relation to competitors. Hambrick 1984

suggests dividing the sample into separate high andlow performance sub-samples in this manner as apractical analytical technique for strategy research.The scores for the performance measures of marketshare and sales growth are combined and used toidentify 51 high performers and 50 low performers.The questions that measure these items use a Likert

Ž .scale ranging from 1 to 7 Appendix A . The scoresfor the performance measures of market share andsales growth are added to create a scale that rangesfrom a low of 2 to a high of 14. Based on thiscomposite score, companies are separated into twogroups: low performers having performance valuesof nine or less and high performers having perfor-mance values of 10 or more.

In addition to the perceptual performance mea-sures discussed above, we also asked participatingfirms for objective performance data on earningsgrowth, although fewer than half of the firms in thesample responded to this questions because the ob-jective data are more sensitive for managers con-cerned about revealing confidential data. Using a

Ž .method described by Vickery et al. 1993 , the corre-lation between perceptual and objective measures isused to help validate subjective measures. The objec-tive measure of earnings growth correlates stronglyŽ .significant at less than 0.001 with the categorical

variable constructed on the basis of composite per-ceptual performance, thus buttressing the perceptualmeasures.

3.3. Reliability

Inter-item analysis is used to check environment,competitive strategy, and manufacturing strategy

Žscales for internal consistency or reliability Nunn-.ally and Bernstein, 1994 . Specifically, Cronbach’s

coefficient alpha is calculated for each scale, asŽ .recommended by Flynn et al. 1990 . Cronbach’s

coefficient alphas, reported in the Table 1, are allacceptable for organizational research according tothe criteria established by Van de Ven and FerryŽ . Ž .1978 pp. 78–81 . Table 1 also contains a correla-tion matrix for the scales described above.

4. Path analytic methods

Causal models have been used in both competi-tive and manufacturing strategy to link environmen-tal constructs to strategy and performance. Swami-

Ž .dass and Newell 1987 showed environment as aprecursor to choice of manufacturing strategy in theirpath model linking perceived environment, opera-tions strategy process, operations strategy content,

Ž .and firm performance. Ward et al. 1995 upheld the


contention that environment is causally linked tooperations strategy choice for high performing firms.

Ž .Gupta and Lonial 1998 use a path model to linkbusiness and manufacturing strategies. We use a pathmodel to link environment, competitive strategy,manufacturing strategy and performance.

Although the relationships among environment,competitive strategy and manufacturing strategy areoften viewed as reciprocal, this study focuses on asingle time period in which competitive strategy canbe seen as a predecessor of manufacturing strategyand environment is a predecessor to both. Thesecausal relationships were captured by employing pathmodels estimated using covariance structure model-ing. We recognize the limitations of using a staticmodel for inferring precedence.

Preliminary path models were first specified toidentify the significant relationships and estimatepath coefficients. Separate models were estimated forboth high and low performance groups. Those co-variance paths that were not significant at greaterthan 0.10 were eliminated to better estimate themodel and provide closer fit. Model fit statisticswere used to test Hypothesis 1, which requires as-sessing model fit for high and low performers. Sig-nificance tests on coefficients for these same modelswere used to evaluate hypotheses 1a through 1c,which concern the existence of relationships betweenenvironmental dynamism and competitive strategy,between competitive strategy and manufacturingstrategy, and between manufacturing strategy andperformance, respectively. Covariance of error termswere specified for all manifest variables to accountfor covariance among factors. Initial path coeffi-cients achieved for the high performer sample areshown in Table 2.

As described in the findings, below, good modelfit was achieved for high performers but not for lowperformers. Therefore, only the high performancesample was retained for subsequent analysis requiredto test hypotheses 2 and 3. To test hypothesis 2, thatenvironmental dynamism directly influences manu-facturing strategy, an additional model was estimated

Žthat provides each manufacturing strategy competi-.tive priority with both a direct link to environment

and an indirect link through competitive strategy.The significance of these path coefficients were usedto test the existence of the causal linkages.

Table 2Path coefficients and significance levels for the initial model

Path Path coefficient t Value

Environmental dynamism toUMarketing differentiation 0.44 3.56

Price y0.08 y0.56Marketing differentiation to

qLow cost 0.25 1.78UUQuality 0.67 4.16qDelivery 0.30 2.16qFlexibility 0.30 2.13

Price toLow cost 0.24 1.59Quality y0.20 y1.13Delivery y0.02 y0.13Flexibility y0.04 y0.29

To PerformanceLow cost y0.04 y0.25Quality 0.30 1.60Delivery 0.14 0.82Flexibility y0.03 y0.13

UPath coefficient significant at -0.05.

UUPath coefficient significant at -0.01.

q Path coefficient significant at -0.10.

Similarly, to test hypothesis 3, that competitivestrategy directly influences business performancestrategy, an additional model was estimated thatprovides a path from each competitive strategy bothdirectly to performance and indirectly through manu-facturing strategy. Again, the significance of thesepath coefficients were used to test the existence ofthe causal linkages.

5. Findings

Fig. 2 shows the significance of paths and pathcoefficients that result from estimating the initialmodel and reducing the model to only paths signifi-cant at 0.10 for high performers. Because no singlemeasure of fit adequately describes covariance struc-

Ž .ture models Bollen and Long, 1993 , Fig. 2 includesthree model fit statistics; Root Mean Square Error of

Ž .Approximation RMSEA ; Bentler–Bonnet normedfit index; and a non-normed fit index. RMSEA esti-mates the model’s fit to the true population parame-ters taking the number of parameters into considera-tion. Probabilities are calculated that the RMSEA


Fig. 2. High performer model.

produces a close fit, i.e., the population discrepancyvalue approaches zero with significance at 0.05, andthe probability of a perfect fit at 0.01. Both the

Bentler–Bonnet normed fit index and the non-normedfit index, which adjusts for degrees of freedom,reflect a comparison with fit indices achieved using a

Fig. 3. Direct environmental effects model.


Ž .restrictive, baseline model Bollen, 1989 . We spec-ify a baseline model representing covariance amongthe competitive strategy variables and between thecompetitive priorities but without paths. For bothBentler–Bonnet and the non-normed fit indices, val-ues close to 1 indicate good fit.

The initial model shows the probability of a per-fect fit of 0.947 and a close fit at 0.969, a normed fitindex of 0.892 and a non-normed index of 1.27. Allfit indices indicate a good fit of the model for highperformers. This indication of a good fit substanti-ates the overall model of the relationship of environ-ment, competitive strategy and manufacturing strat-egy for high performing firms.

The same model was specified for low performerswith dramatically different results. The fit statisticsdo not indicate a good fit of the model when datafrom low performers is used. The low performermodel shows the probability of a perfect fit of 0.169and a close fit at 0.275, a normed fit index of 0.673.and a non-normed index of 0.859. The indication ofa good fit for the high performer model and poor fit

for the low performing model supports hypothesis 1,that the predominant manufacturing strategy modelfits high performers and not poor performers. Be-cause the low performer model fits poorly, subse-quent discussion is restricted to the high performancesubsample.

The existence of significant paths from environ-ment to differentiation and from competitive strategyvariables to manufacturing strategy variables pro-vides evidence of a causal relationship existing be-tween environment and competitive strategy and be-tween competitive strategy and manufacturing strat-egy. Similarly, a significant path between one of the

Ž .manufacturing strategy dimensions quality and per-formance also indicates the predicted relationshipbetween manufacturing strategy and performance.These findings support hypothesis 1a, that environ-mental dynamism affects competitive strategy choice;hypothesis 1b, that competitive strategy influencesthe selection of an appropriate manufacturing strat-egy; and hypothesis 1c, that manufacturing strategyis related to performance.

Fig. 4. Direct competitive strategy effects model.


Hypothesis 1a requires that at least one significantpath exists between environment and competitivestrategy. The path from environmental dynamism todifferentiation is significant at the 0.05 level. Simi-larly, hypothesis 1b requires that at least one signifi-cant path exists between competitive strategy andmanufacturing strategy. The competitive strategy ofdifferentiation is positively linked to the quality andflexibility dimensions of manufacturing strategy at

Žthe 0.05 significance level with the other paths.significant at 0.10 . Hypothesis 1c requires that a

significant path exist between manufacturing strategyand performance. The path from the manufacturingstrategy dimension of quality to business perfor-mance is significant at the 0.05 level.

To test hypothesis 2, that a direct relationshipexists between environmental dynamism and manu-facturing strategy, the initial high performer modelwas re-estimated with the addition of paths from theenvironment to manufacturing strategy constructs.Support for hypothesis 2 requires that at least one ofthese new paths be significant. This model and re-sults are summarized in Fig. 3.

Fig. 3 shows that the four paths from environmentto manufacturing competitive priorities are not sig-nificant at 0.10. Thus, hypothesis 2, that environmen-tal factors directly influence manufacturing strategyamong high performance manufactures, is NOT sup-ported. Because these paths are not significant, thereis no compelling evidence of a direct relationshipbetween environment and manufacturing strategywhen competitive strategy is also considered.

Finally, hypothesis 3 requires a significant pathbetween either of the competitive strategy variatesand business performance. Fig. 4 shows the highperformer model respecified to include direct pathsfrom each of the competitive strategy dimensions,price and differentiation, to business performance.Because neither of these paths are significant at 0.10,the data do not provide evidence of a direct competi-tive strategy–business performance relationship.

In summary, the data support hypotheses 1, 1a, 1band 1c and thus the conceptual model of manufactur-ing strategy that has been predominant in the litera-ture. However, the data do not support hypotheses 2and 3. The implications of these findings are dis-cussed below.

6. Discussion

In essence, this research suggests four notablefindings. First, long-standing conceptual argumentslinking environment, competitive strategy, manufac-turing strategy, and performance are upheld empiri-cally among high performance firms. Second, themodel does not fit firms that report relatively poorbusiness performance. Third, we find no direct linkbetween environmental dynamism and manufactur-ing strategy; rather we find that this relationship ismediated by competitive strategy. Fourth, the data donot support a direct relationship between competitivestrategy and business performance. The data suggestthat the relationship between competitive strategyand performance is mediated by manufacturing strat-egy. We elaborate on each of these findings in turn.

6.1. Empirical support for the conceptual model

The tests of the causal model shown in Fig. 2confirm widely held beliefs about the role of manu-facturing strategy in context. Although this findingonly confirms what many scholars already hold to betrue, it is worth highlighting because such confirma-tion has not been reported in the literature previ-ously. Although a number of empirical studies haveaddressed pairs of constructs in the environment–competitive strategy–manufacturing strategy nexus,a review of the literature has not revealed an instancewhere all three appear in a model simultaneously. Inaddition, we address the performance implicationssuggested by the model.

The data analysis also reveals information aboutthe specific nature of the relationships that exist inour sample of high performance manufacturing firms.Recall that we use two predominant competitivestrategy dimensions, differentiation and price. The-ory and empirical evidence in competitive strategysuggests that differentiation strategies are more ef-

Žfective in dynamic environments Miller, 1986, 1988;.Bourgeois and Eisenhardt, 1988 . Our own findings

confirm this position, the path between environmen-tal dynamism and a differentiation strategy is signifi-cant and positive while the path between environ-


mental dynamism and a low price competitive strat-Ž .egy is not significant see Fig. 2 .

From the perspective of operations management,the paths between each of the competitive strategiesand the manufacturing strategy dimensions are ofgreat interest. A competitive strategy of differentia-tion is linked with each of the manufacturing strat-egy variables. The coefficient of the paths betweendifferentiation and quality and between differentia-tion and flexibility are each significant at less than0.05. Links between differentiation and the other twomanufacturing strategy dimensions, low cost anddelivery, are significant at 0.10. This finding sug-gests that successful differentiators pursue a portfolioof manufacturing capabilities to make their offeringdistinctive in the marketplace. The fact that qualityshows the strongest link with differentiation is con-

Žsistent with the literature e.g., Garvin, 1987;.Williams et al., 1995 . As expected, the model re-

flects a relationship between a low price competitivestrategy and a low cost manufacturing strategy, a

Žfinding suggested by conceptual literature e.g., Ward.et al., 1996 and common sense. The price–low cost

path is significant at 0.10.The strong link between quality and business

performance is also notable. This finding is consis-tent with both the vast body of TQM research thatsuggests that a quality emphasis is primary. It is alsoconsistent with a number of empirical studies thatsuggest a positive relationship between quality and

Žvarious measures of business performance e.g.,Buzzell and Gale, 1987; Flynn et al., 1995a; Williams

.et al., 1995 . Findings reported by Narasimhan andŽ .Jayaram 1998 suggest programs that are an-

tecedents to making progress in achieving competi-tive priorities, including quality.

The importance of the close coupling betweencompetitive and manufacturing strategies among highperformance manufacturers raises interesting ques-tions about how such coupling can be accomplished

Ž .Hill 1994 provides one methodology for achievingsuch a coupling and also points out many potential

Ž .pitfalls in the process. Adam and Swamidass 1989and others point out that manufacturing strategyprocess research has been neglected relative to con-

Ž .tent research. The content research findings re-ported here underline the importance of process re-search for developing an understanding how firms

establish close linkages between competitive andoperations strategy without adopting bureaucraticstrictures that impede responsiveness.

6.1.1. Poor fit for poor performersA number of authors have suggested the conse-

quences of not adhering to a manufacturing strategymodel that ties business and manufacturing strategy:

Ž .poor performance e.g., Skinner, 1969; Hill, 1994 .Our findings of poor model fit for poor performersare consistent with the admonitions of these andother influential thinkers in our field. Our findingsalso suggest the practical advice for empirical re-searchers in our field, separate consideration of high

Ž .and low performers. Hambrick 1984 specificallyrecommends splitting a sample based on perfor-mance to capture the different behaviors in strategyresearch using a configurational approach. Other ap-proaches suggest different means for achieving sepa-ration but the idea that poor performers behaviormay be fundamentally different from good perform-ers is worthwhile.

6.2. EnÕironmental dynamism and manufacturingstrategy

At first glance, the finding of no direct relation-ship between environmental dynamism and manufac-turing strategy appears at variance with the empiricalfindings reported by Van Dierdonck and MillerŽ . Ž .1980 , Swamidass and Newell 1987 and Ward et

Ž .al. 1995 . The apparent inconsistency is easily ex-plained, however, by the fact that none of the studiesnoted above considered environment and competitivestrategy simultaneously. Our findings indicate thatcompetitive strategy mediates the effects of environ-mental dynamism on manufacturing strategy in highperformance firms.

Testing for mediation is usually done in a pathanalytic framework similar to the one used in thisresearch. The significant paths between environmen-tal dynamism and competitive strategy and betweencompetitive strategy and manufacturing strategy cast

Žcompetitive strategy as a mediator Venkatraman,.1989 . The fact that there is not a direct path be-

tween environmental dynamism and manufacturingŽstrategy provides stronger evidence of mediation i.e.,

evidence of complete mediation, Blalock, 1969;


.Venkatraman, 1989 . The mediating effect of com-petitive strategy suggests that environmental dy-namism has an important influence on manufacturingstrategy but that influence is articulated through andmodified by competitive strategy.

The research implication of competitive strategymediating the effects of environment on manufactur-ing strategy is clear. A model of manufacturingstrategy must include both environmental and com-petitive strategy variables to capture the context ofmanufacturing strategy accurately. Previous empiri-cal research in manufacturing strategy generally ex-cludes from consideration either environment orcompetitive strategy. Our results suggest that over-looking either environment or competitive strategymay miscast the true relationships. Therefore, it isimportant to assess manufacturing strategy in thecontext of both environment and competitive strat-egy. In short, the data suggest that for high perfor-mance manufacturers, reactions to environmentalconditions are effected through competitive strategy.This only underscores the importance of a closecoupling of competitive and manufacturing strategiesand, again, implies that learning how to make effec-tive links between competitive and manufacturingstrategy is critical.

6.3. CompetitiÕe strategy and performance

Our analysis suggests that the relationship be-tween competitive strategy and business performanceis mediated by manufacturing strategy. More specifi-cally, the quality dimension of manufacturing strat-egy appears to mediate the differentiation strategy–business performance connection. This finding im-plies that a differentiation strategy works when it issupported by manufacturing capability, i.e., quality.

This implication is important because it suggests thatperformance improvements resulting from competi-tive strategy initiatives are manifested in their imple-mentation via manufacturing capabilities, specificallyquality. By using more precise instruments to mea-sure competitive strategy, future research may dis-cern that other dimensions of manufacturing strategyalso serve to define the performance effects of com-petitive strategy.

The emerging paradigm of manufacturing strategythat appears in the literature suggests tight constella-tions of environmental factors and strategies which

Žlead to superior capabilities and performance e.g.,.Miller and Roth, 1994; Hayes and Pisano, 1996 .

Our study supports this view of manufacturing strat-egy. In addition, we test and support a model ofmanufacturing strategy that is predominant in theconceptual literature for high performers but showthat this model does not fit for low performers. Thisresearch also shows that competitive strategy is amediator between environment and manufacturingstrategy for high performing firms. Perhaps mostnotably, the findings provide empirical evidence thatmanufacturing and competitive strategies are inextri-cably linked in high performance firms. This sug-gests that research into the processes that companiesuse to achieve those links is of great importance formoving forward our knowledge of manufacturingstrategy.

Acknowledgements

We thank the Ohio State University’s Center forExcellence in Manufacturing Management for finan-cial support. Errors remain the responsibility of theauthors.

Appendix A. Items used for developing scales

EnÕironmental dynamismIndicate the rate of change for the following

Slow Rapid NrOØThe rate at which products and services become outdated 1 2 3 4 5 6 7 9ØThe rate of innovation of new products or services 1 2 3 4 5 6 7 9ØThe rate of innovation of new operating processes 1 2 3 4 5 6 7 9ØThe tastes and preferences of customers in your industry 1 2 3 4 5 6 7 9


CompetitiÕe strategyRate the following competitive methods on how important they are in meeting your business strategy

No Very Absolutely NrOimportance important critical

PriceØOperating efficiency 1 2 3 4 5 6 7 9ØCompetitive pricing 1 2 3 4 5 6 7 9ØProcurement of raw materials 1 2 3 4 5 6 7 9ØReducing product costs 1 2 3 4 5 6 7 9ØMinimize outside financing 1 2 3 4 5 6 7 9ØDecreasing the number of 1 2 3 4 5 6 7 9

product featuresDifferentiationØNew product development 1 2 3 4 5 6 7 9ØBrand identification 1 2 3 4 5 6 7 9Ø Innovation in marketing 1 2 3 4 5 6 7 9

techniques and methodsØControl of distribution channels 1 2 3 4 5 6 7 9ØAdvertising 1 2 3 4 5 6 7 9

Manufacturing strategyIndicate the degree of emphasis which your manufacturing plant places on the following activities

No Moderate Extreme NrOemphasis emphasis emphasis

FlexibilityØLead-time reduction 1 2 3 4 5 6 7 9ØSetup time reduction 1 2 3 4 5 6 7 9ØAbility to change priorities of 1 2 3 4 5 6 7 9

jobs on the shop floorØAbility to change machine 1 2 3 4 5 6 7 9

assignments of jobs onthe shop floor

QualityØStatistical process control 1 2 3 4 5 6 7 9ØReal-time process control systems 1 2 3 4 5 6 7 9ØUpdating process equipment 1 2 3 4 5 6 7 9ØDeveloping new processes 1 2 3 4 5 6 7 9

for new productsØDeveloping new processes 1 2 3 4 5 6 7 9

for old productsDeliveryØProvide fast deliveries 1 2 3 4 5 6 7 9ØMeet delivery promises 1 2 3 4 5 6 7 9CostØReduce inventory 1 2 3 4 5 6 7 9Ø Increase capacity utilization 1 2 3 4 5 6 7 9Ø Increase equipment utilization 1 2 3 4 5 6 7 9ØReduce production costs 1 2 3 4 5 6 7 9


PerformanceCompared to your competitors, indicate your position on the following dimensions

Significantly SignificantlyLower Equal Higher NrO

ØMarket share 1 2 3 4 5 6 7 9ØSales growth 1 2 3 4 5 6 7 9

References

Adam, E.E. Jr., Swamidass, P.M., 1989. Assessing operationsmanagement from a strategic perspective. Journal of Manage-

Ž .ment 15 2 , 181–203.Anderson, J.C., Cleveland, G., Schroeder, R.G., 1989. Operations

strategy: a literature review. Journal of Operations Manage-Ž .ment 8 2 , 133–158.

Blalock Jr., H.M., 1969. Theory Construction: From Verbal toMathematical Formulations. Prentice-Hall, Englewood Cliffs,NJ.

Bollen, K.A., 1989. Structural Equations with Latent Variables.Wiley, New York.

Ž .Bollen, K.A., Long, J.S. Eds. , 1993. Testing Structural EquationModels. Sage, Newbury Park.

Bourgeois, L.J., Eisenhardt, K.M., 1988. Strategic decision pro-cesses in high velocity environments: four cases in the micro-

Ž .computer industry. Management Sciences 34 7 , 816–834.Buzzell, R.D., Gale, B.T., 1987. The PIMS principles. The Free

Press, New York, NY.Collins, R.S., Cordon, C., Julien, D., 1998. An empirical test of

the rigid flexibility model. Journal of Operations Management16, 133–146.

Ž .Dess, G.G., Davis, P.S., 1984. Porter’s 1980 generic strategiesas determinants of strategic group membership and organiza-

Ž .tional performance. Academy of Management Journal 27 3 ,467–488.

Doty, D.H., Glick, W.H., Huber, G.P., 1993. Multidimensional fit,equifinality, and organizational effectiveness: a test of twoconfigurational theories. Academy of Management Journal 36Ž .6 , 1196–1250.

Ferdows, K., DeMeyer, A., 1990. Lasting improvements in manu-facturing performance: in search of a new theory. Journal of

Ž .Operations Management 9 2 , 168–184.Flynn, B.B., Sakakibara, S., Schroeder, R., Bates, K., Flynn, J.,

1990. Empirical research methods in operations management.Ž .Journal of Operations Management 9 2 , 250–284.

Flynn, B.B., Schroeder, R., Sakakibara, S., 1994. A frameworkfor quality management and an associated measurement instru-ment. Journal of Operations Management 11, 339–366.

Flynn, B.B., Sakakibara, S., Schroeder, R., 1995a. Relationshipbetween JIT and TQM: practices and performance. Academy

Ž .of Management Journal 38 5 , 1325–1360.Flynn, B.B., Schroeder, R., Sakakibara, S., 1995b. The impact of

quality management practices on performance and competitiveŽ .advantage. Decision Sciences 26 5 , 659–691.

Garvin, D.A., 1987. Competing on the eight dimensions of qual-Ž .ity. Harvard Business Review 65 6 , 101–109.

Gerwin, D., 1993. Manufacturing flexibility: a strategic perspec-tive. Management Science 39, 395–410.

Gupta, Y.P., Lonial, S.C., 1998. Exploring linkages betweenmanufacturing strategy, business strategy, and organizational

Ž .strategy. Production and Operations Management 7 3 , 243–264.

Hambrick, D.C., 1984. Taxonomic approaches to studying strat-egy. Journal of Management 10, 27–41.

Hatten, K., Schendel, D., Cooper, A., 1978. A strategic model ofthe U.S. brewing industry: 1952–1971. Academy of Manage-ment Journal 21, 592–610.

Hayes, R.H., Wheelwright, S.C., Clark, K., 1988. Dynamic manu-facturing. The Free Press, New York.

Hayes, R.H., Pisano, G.P., 1996. Manufacturing strategy: at theintersection of two paradigm shifts. Production and Operations

Ž .Management 5 1 , 25–41.Hayes, R.H., Wheelwright, S.C., 1984. Restoring our Competitive

Edge: Competing Through Manufacturing. Wiley, New York.Hill, T.J., 1994. Manufacturing Strategy: Text and Cases, 2nd edn.

Irwin, Burr Ridge, IL.Hofer, C.W., 1975. Toward a contingency theory of business

strategy. Academy of Management Journal 18, 784–810.Keats, B.W., Hitt, M.A., 1988. A causal model of linkages among

environmental dimensions, macro-organizational character-istics and performance. Academy of Management Journal 31Ž .3 , 570–598.

Kim, L., Lim, Y., 1988. Environment, generic strategies, andperformance in a rapidly developing location: a taxonomic

Ž .approach. Academy of Management Journal 31 4 , 802–827.Kotha, S., Vadlamani, B.L., 1995. Assessing generic strategies: an

empirical investigation of two competing typologies in dis-crete manufacturing industries. Strategic Management Journal16, 75–83.

Leong, G.K., Snyder, D., Ward, P.T., 1990. Research in theprocess and content of manufacturing strategy. OMEGA 18Ž .2 , 109–122.

Miller, D., 1986. Configurations of strategy and structure: towardsŽ .a synthesis. Strategic Management Journal 7 3 , 233–249.

Miller, D., 1987. The structural and environmental correlates ofŽ .business strategy. Strategic Management Journal 8 1 , 55–76.

Miller, D., 1988. Relating Porter’s business strategies to environ-ment and structure: analysis and performance implications.

Ž .Academy of Management Journal 31 2 , 280–308.Miller, D., Friesen, P.H., 1983. Strategy-making and environment:

Ž .the third link. Strategic Management Journal 4 3 , 221–235.


Miller, J.G., Roth, A.V., 1994. A taxonomy of manufacturingŽ .strategies. Management Science 40 3 , 285–304.

Narasimhan, R., Jayaram, J., 1998. An empirical investigation ofthe antecedents and consequences of manufacturing goalachievement in North American, European, and Pan Pacificfirms. Journal of Operations Management 16, 159–176.

Noble, M.A., 1995. Manufacturing strategy: testing the cumula-tive model in a multiple country context. Decision Sciences 26Ž .5 , 693–720.

Nunnally, J.C., Bernstein, I.H., 1994. Psychometric Theory, 3rdedn. McGraw-Hill, New York.

Porter, M.E., 1980. Competitive Strategy. The Free Press, NewYork.

Skinner, W., 1969. Manufacturing — missing link in corporateŽ .strategy. Harvard Business Review 47 3 , 136–145.

Swamidass, P.M., Newell, W.T., 1987. Manufacturing strategy,environmental uncertainty and performance: a path analytic

Ž .model. Management Science 33 4 , 509–524.Swink, M., Way, M.H., 1995. Manufacturing strategy: proposi-

tions, current research, renewed directions. International Jour-Ž .nal of Operations and Production Management 15 7 , 4–26.

Van de Ven, A.H., Drazin, R., 1985. The concept of fit inŽ .contingency theory. In: Cummins, L., Staw, B.M. Eds. ,

Research in Organizational Behavior, Vol. 7. JAI Press,Greenwich, CT, pp. 333–365.

Van de Ven, A.H., Ferry, D., 1978. Measurement and Assessmentof Organizations. Wiley, New York.

Van Dierdonck, R., Miller, J.G., 1980. Designing productionplanning and control systems. Journal of Operations Manage-

Ž .ment 1 1 , 37–46.Venkatraman, N., 1989. The concept of fit in strategy research:

towards verbal and statistical correspondence. Academy ofŽ .Management Journal 14 3 , 423–444.

Vickery, S.K., Droge, C., Markland, R.R., 1993. Production com-petence and business strategy: do they affect business perfor-

Ž .mance. Decision Sciences 24 4 , 435–456.Ward, P.T., Duray, R., Leong, G.K., Sum, C., 1995. Business

environment, operations strategy, and performance: an empiri-cal study of Singapore manufacturers. Journal of Operations

Ž .Management 13 2 , 99–115.Ward, P.T., Bickford, D.J., Leong, G.K., 1996. Configurations of

manufacturing strategy, business strategy, environment, andstructure. Journal of Management 22, 597–626.

Wheelwright, S.C., 1984. Manufacturing strategy: defining theŽ .missing link. Strategic Management Journal 5 1 , 77–87.

Williams, F.P., D’Souza, D.E., Rosenfeldt, M.E., Kassaee, M.,1995. Manufacturing strategy, business strategy and firm per-formance in a mature industry. Journal of Operations Manage-ment 13, 19–33.

Youndt, M.A., Snell, S.A., Dean, J.W. Jr., Lepak, D.P., 1996.Human resource management, manufacturing strategy and firmperformance. Academy of Management Journal 39, 836–866.

Documents

Manufacturing strategy in context: environment, competitive strategy and manufacturing strategy