total quality management as competitive advantage

Strategic Management Journal, Vol. 16, 15-37 (1995)

TOTAL QUALITY MANAGEMENT AS COMPETITIVEADVANTAGE: A REVIEW AND EMPIRICAL STUDYTHOMAS C. POWELLManagement Department, Bryant College, Smithfield, Rhode Island, U.S.A.

Total Quality Management (TQM) has become, according to one source, 'as pervasive apart of business thinking as quarterly financial results,' and yet TQM's role as a strategicresource remains virtually unexamined in strategic management research. Drawing on theresource approach and other theoretical perspectives, this article examines TQM as apotential source of sustainable competitive advantage, reviews existing empirical evidence,and reports findings from a new empirical study of TQM's performance consequences. Thefindings suggest that most features generally associated with TQM—such as quality training,process improvement, and benchmarking—do not generally produce advantage, but thatcertain tacit, behavioral, imperfectly imitable features—such as open culture, employeeempowerment, and executive commitment—can produce advantage. The author concludesthat these tacit resources, and not TQM tools and techniques, drive TQM success, and thatorganizations that acquire them can outperform competitors with or without the accompanyingTQM ideology.

INTRODUCTION

Over the past 10 years, the emphasis in strategicmanagement thinking has shifted away fromindustry structure and competitive positioning,and toward internal, firm-specific, 'within stra-tegic group' factors (Cool and Schendel, 1988)such as culture (Barney, 1968a; Fiol, 1991),capabilities (Lawless, Bergh, and Wilsted, 1989;Stalk, Evans and Shulman, 1992), administrativeskills (Powell, 1992), reputation (Weigelt andCamerer, 1988), know-how (Hall, 1992), learning(Senge, 1990; Garvin, 1993), process improve-ment (Stalk and Hout, 1990), and organizationalclimate (Hanson and Wernerfelt, 1989), Theresource theory of the firm has accelerated thisshift, asserting that economic rents may stemfrom any strategic factor—internal, external.

Key words: TQM, competitive advantage, resourcetheory

economic, behavioral, tangible, or intangible—that meets the tests of value, scarcity, andimperfect imitability (Wernerfelt, 1984; Barney,1986b; 1991; Peteraf, 1993),

By no coincidence, the past 10 years have alsowitnessed the remarkable spread of Total QualityManagement (TQM). A recent Industry Weekarticle (Benson, 1993: 48) claimed that, 'In 10short years, TQM has become as pervasive apart of business thinking as quarterly financialresults', and a recent Arthur D, Little studyreported that 93 percent of America's largest500 firms had adopted TQM in some form(Arthur D, Little, 1992), Analysts have creditedTQM with leading Japan to global economicprominence in the postwar years (Grayson andQ'Dell, 1988; Imai, 1986) and, more recently,with restoring America's economic competi-tiveness (Juran, 1993). To reward exemplaryTQM initiatives, the U.S. Department of Com-merce instituted the Malcolm Baldrige QualityAward in 1987, and TQM has brought public

CCC 0143-2095/95/010015-23© 1995 by John Wiley & Sons, Ltd,

Received 25 October 1993Final revision received 21 February 1994

16 T.C. Powell

recognition to such gurus as Joseph Juran, PhilipCrosby, and the late W, Edwards Deming (afterwhom Japan's esteemed Deming Award isnamed),

TQM is an integrated management philosophyand set of practices that emphasizes, amongother things, continuous improvement, meetingcustomers' requirements, reducing rework, long-range thinking, increased employee involvementand teamwork, process redesign, competitivebenchmarking, team-based problem-solving, con-stant measurement of results, and closer relation-ships with suppliers (Ross, 1993), Its adherentsclaim that managers can implement TQM in anyorganization—manufacturing, service, nonprofit,or government—and that it generates improvedproducts and services, reduced costs, moresatisfied customers and employees, and improvedbottom line financial performance (Walton,1986).

The latter claim is controversial. Althoughmany adherents openly praise TQM, others haveidentified significant costs and implementationobstacles. Critics have suggested, for example,that TQM entails excessive retraining costs,consumes inordinate amounts of managementtime, increases paperwork and formality,demands unrealistic employee commitment lev-els, emphasizes process over results, and fails toaddress the needs of small firms, service firms,or nonprofits (Naj, 1993; Fuchsberg, 1992a;1993b; Schaffer and Thomson, 1992), Indeed,the Wallace Company, a Houston oil-supplyfirm, filed Chapter 11 bankruptcy soon afterwinning the Baldrige Award (Hill, 1993), andTQM exemplar and Deming Award winnerFlorida Power and Light virtually eliminated itsprogram over employee complaints of excessivepaperwork. Moreover, empirical studies havenot shown that TQM firms consistently outper-form non-TQM firms (Mathews, 1992; Fuchsberg,1993a),

Nevertheless, TQM has become an irrepress-ible, globally pervasive strategic force in today'sindustrial economy. And, because TQM requiresfirms to coordinate a wide range of behavioral,tacit, intangible resources, its disseminationstands as both support and a challenge to thenew emphasis on firm-specific resources instrategic management research,

TQM's impact on strategic managementresearch and practice remains unclear and under-

examined, and the existing empirical studies ofTQM performance—intended to help managersimplement TQM more effectively—lack rigorand theoretical support. This article attempts toredress these problems, examining the impli-cations of TQM for strategic managementresearch and practice, evaluating TQM from theresource and other theoretical perspectives, andpresenting an empirical test of TQM performanceimpacts.

TQM AND FIRM PERFORMANCE

Origins and elements of TQM

TQM's origins can be traced to 1949, when theUnion of Japanese Scientists and Engineers(JUSE) formed a committee of scholars, engi-neers, and government officials devoted toimproving Japanese productivity, and enhancingtheir postwar quality of life. Influenced byDeming and Juran, the committee developed acourse on statistical quality control for Japaneseengineers, followed by extensive statistical train-ing and the widespread dissemination of theDeming philosophy among Japanese manufac-turers (Walton, 1986),

In Japan, TQM produced such managerialinnovations as quality circles, equity circles,supplier partnerships, cellular manufacturing,just-in-time production, and hoshin planning(Ishikawa, 1985; Akao, 1991), However, asquality control programs became more widelyimplemented and sophisticated, it became clearthat some aspects of the TQM philosophy couldalso be applied to nonmanufacturing functionssuch as product development, purchasing, andbilling, with potential applications in serviceorganizations and nonprofits,

American firms began to take serious notice'of TQM around 1980, when some U,S, policyobservers argued that Japanese manufacturingquality had equaled or exceeded U,S, standards,and warned that Japanese productivity wouldsoon surpass that of American firms (e,g,,Hayes and Abernathy, 1980), Productivity trendssupported these assertions, leading some opinionleaders to predict that—barring a radical changein American management practices—Japan andother Asian countries would soon dominateworld trade and manufacturing, relegating theU,S, to second-tier economic status (e,g,. Gray-

TQM as Competitive Advantage 17

son and O'Dell, 1988). In particular, theseanalysts decried traditional American managerialpractices such as elitist leadership, autocraticauthority structures, short-term thinking, finan-cial orientation, lack of innovation, decliningproduct quality, adversarial supplier relationships(including employees), inadequate training, and,in general, living off past successes (Hayes andAbernathy, 1980; Pascale, 1981; Grayson andO'Dell, 1988; Jacobs, 1991).

Some high-profile American firms—such asFord, Xerox, and Motorola—were easily con-vinced, having already lost market share to moreefficient, higher quality Japanese producers.These firms, under the guidance of Deming andother quality consultants, benchmarked Japanesepractices and were among the first AmericanTQM adopters. Based on their widely-publicizedsuccesses, other large manufacturers soon jumpedaboard, and by the end of the 1980s, a significantproportion of large U.S. manufacturers hadadopted TQM (Arthur D. Little, 1992). By thattime, many large service firms had also expressedinterest, and some—due in part to pressures fromcustomers that employed TQM—had adopted fullTQM initiatives.

Table 1 lists the major TQM features pro-moted by Deming, Juran, and Crosby, alongwith the categories used to evaluate BaldrigeAward candidates. Although different TQMproponents emphasize different features (e.g.,Deming focuses on statistical quality measures),an exhaustive review and integration of theTQM literature suggests that complete TQMprograms tend to share the 12 factors shownin Table 2.

The economic value of TQM

TQM has disseminated widely among Fortune1000 firms, presumably because managers believeTQM improves performance. However, both theanecdotal evidence and the empirical studiessuggest considerable variability in TQM's per-formance impacts, ranging from unprecedentedsuccesses to bankruptcy and abandonment ofTQM. Can TQM act as a source of sustainedcompetitive advantage? If not, why is TQMdisseminating so rapidly? If so, why the mixedresults and high-profile failures?

Resource theory provides a useful perspectiveon these issues, beginning with the notion of

resource heterogeneity, i.e., that different firmshold different resource portfolios, and that thesedifferences produce variability in performanceacross firms (Wernerfelt, 1984; Barney, 1986b;Peteraf, 1993). Although firms may attempt toimitate resources held by successful competitors,or at least to replicate their benefits, resourcebundles remain heterogeneous due to imperfectimitability, created by 'isolating mechanisms'(Rumelt, 1984) such as: (1) time compressiondiseconomies—the resource may require long-term accumulation before attaining value (e.g.,learning, experience, or proficiency in a skill);(2) historical uniqueness (first moveradvantages)—the resource may have been orig-inally acquired under unique, nonreplicable con-ditions; (3) connectedness of resources—a firmmay acquire a competitor's valuable resourceonly to find that its success depends on somecomplementary resource that the firm cannotacquire; (4) causal ambiguity—firms may beunable to determine the link between anotherfirm's resources and its success; and (5) socialcomplexity—a firm's success may result fromsocial phenomena too complex for managers tounderstand or manage (Lieberman and Montgom-ery, 1988; Dierickx and Cool, 1989; Barney,1991).

Under the resource view, success derives fromeconomically valuable resources that other firmscannot imitate, and for which no equivalentsubstitute exists. Is TQM such a resource?According to TQM advocates, TQM does pro-duce value, through a variety of benefits:improved understanding of customers' needs;improved customer satisfaction; improvedinternal communication; better problem-solving;greater employee commitment and motivation;stronger relationships with suppliers; fewer errors;and reduced waste (Juran, 1988; Schmidt andFinnigan, 1992; Spechler, 1991). But the evidencealso suggests that some employees resist or evensubvert TQM, finding it ideological or faddish.Furthermore, TQM entails substantial timeinvestments from managers, it is expensive(especially for training and meetings), it rarelyproduces short-term results, it demands intenseCEQ commitment, and it makes unrealisticassumptions about most organizations' capacitiesto transform their cultures (Bleakley, 1993; Naj,1993; Fuchsberg, 1992, 1993a, 1993b; Mathews,1992).

18 T.C. Powell

Table 1. Popular perspectives on TQM

DEMING'S 14 POINTS'1. Constancy of Purpose2. Adopt the Philosophy3. Don't rely on mass inspection4. Don't award business on price5. Constant improvement6. Training7. Leadership8. Drive out fear9. Break down barriers

10. Eliminate slogans andexhortations

11. Eliminate quotas12. Pride of workmanship13. Education and retraining14. Plan of action

THE JURAN TRILOGY^I. Quality PlanningSet goalsIdentify customers and their needsDevelop products and processes

II. Quality controlEvaluate performanceCompare to goals and adapt

III. Quality improvementEstablish infrastructureIdentify projects and teamsProvide resources and trainingEstablish controls

CROSBY'S 14 QUALITY STEPS'1. Management commitment2. Quality improvement teams3. Quality measurement4. Cost of quality evaluation5. Quality awareness6. Corrective action7. Zero-defects committee8. Supervisor training9. Zero-defects day

10. Goal-setting11. Error cause removal12. Recognition13. Quality councils14. Do it over again

THE 1992 BALDRIGE AWARD1. Leadership (90 points)

1.1 Senior executive1.2 Management for quality1.3 Public responsibility

2.0 Information and analysis (80 points)2.1 Scope and management of quality and

performance data2.2 Competitive comparisons and benchmarks

3.0 Strategic quality planning (60 points)3.1 Strategic quality and planning process3.2 Quality and performance plans

4.0 Human resource development and management(150 points)

4.1 Human resource management4.2 Employee involvement4.3 Employee education and training4.4 Employee performance and recognition4.5 Employee well-being and morale

CRITERIA (1000 points total)'*5.0 Management of process quality (140 points)

5.1 Design and introduction of products andservices

5.2 Process management—production and delivery5.3 Process management—business and support5.4 Supplier quality5.5 Quality assessment

6.0 Quality and operational results (180 points)6.1 Product and service quality6.2 Company operations6.3 Business process and support services6.4 Supplier quality

7.0 Customer focus and satisfaction (300 points)7.1 Customer relationships7.2 Commitment to customers7.3 Customer satisfaction determination7.4 Customer satisfaction results7.5 Customer satisfaction comparisons7.6 Future requirements and expectations

Sources: 'Walton (1986). Juran (1992), 'Crosby (1979). -"George (1992).

The empirical research

Most existing empirical studies conclude thatTQM does produce value. However, most of thestudies were conducted by consulting firms orquality associations with vested interests in theiroutcomes, and most did not conform withgenerally-accepted standards of methodologicalrigor. For example, in 1983 the Union of JapaneseScientists and Engineers published a study ofJapanese companies that won the Deming Prizebetween 1961 and 1980. The study concludedthat these firms had maintained above-averagelong-range performance, as measured by earn-ings, productivity, growth rates, liquidity, and

worker safety. However, the study did notinclude firms that had not won the Deming Prize(one would not expect the experiences of DemingAward winners to represent all firms' experienceswith TQM), it did not report on the progress ofnonTQM firms over the same period, and it didnot control for industry factors that might haveproduced the observed performance differences.

In 1989, The Conference Board, a New Yorkbusiness research group, released a study of thequality practices of large U.S. corporations. Theyreceived 149 responses to 800 surveys, andreported that 111 (74.5%) had quality initiativesin place. Qver 30 percent said that TQM hadimproved their performance, with less than 1


Table 2. The twelve TQM factors

1. Committed leadership: a near-evangelical, unwavering, long-term commitment by topmanagers to the philosophy, usually under a name something like Total QualityManagement, Continuous Improvement (CI), or Quality Improvement (QI).2. Adoption and communication of TQM: using tools like the mission statement, and themesor slogans.3. Closer customer relationships: determining customers' (both inside and outside the firm)requirements, then meeting those requirements no matter what it takes.4. Closer supplier relationships: working closely and cooperatively with suppliers (often sole-sourcing key components), ensuring they provide inputs that conform to customers' end-userequirements.5. Benchmarking: researching and observing best competitive practices.6. Increased training: usually includes TQM principles, team skills, and problem-solving.7. Open organization: lean staff, empowered work teams, open horizontal communications,and a relaxation of traditional hierarchy.8. Employee empowerment: increased employee involvement in design and planning, andgreater autonomy in decision-making.9. Zero-defects mentality: a system in place to spot defects as they occur, rather thanthrough inspection and rework.10. Flexible manufacturing: (applicable only to manufacturers) can include just-in-timeinventory, cellular manufacturing, design for manufacturability (DFM), statistical processcontrol (SPC), and design of experiments (DQE).11. Process improvement: reduced waste and cycle times in all areas through cross-departmental process analysis.12. Measurement: goal-orientation and zeal for data, with constant performancemeasurement, often using statistical methods.

percent reporting performance declines as aresult of TQM. This study did not control forindustry factors, did not include medium-sizedor small firms, and did not track the performanceof comparable non-TQM firms over the sameperiod. In 1989, the Gallup Qrganization surveyed600 senior executives on behalf of The AmericanSociety for Quality Control. The study reportedthat 54 percent of respondents were at least'pleased' with their quality efforts, and half ofthese claimed significant performance impacts.The study focused on large firms and did notcontrol for industry factors.

In 1991, the U.S. Government GeneralAccounting Qffice (GAQ), responding to arequest from the U.S. Congress, produced astudy of the 20 highest-scoring applicants for the1988 and 1989 Baldrige Awards (U.S. GAQ,1991). The GAQ reported that these firms hadachieved better employee relations, improvedproduct quality, lower costs, and improvedcustomer satisfaction. According to the study,however, the methodology did not constitute 'astatistically-rigorous analysis of the companies'performance under quality management' (1991:3); indeed, the study did not control for industry

factors, did not include firms that did not applyfor the Baldrige, and did not report on theprogress of non-TQM firms over the same period.

Several large TQM consulting firms haveproduced in-house quality studies for their clients,but most have released only their results, andnot their methodologies. In 1992, the Authur D.Little Corporation produced an in-house reportbased on a survey of 500 large U.S. firms(Arthur D. Little, 1992). Ninety-three percentof respondents claimed to have some form ofTQM, with 35 percent reporting that theirTQM efforts have had 'significant performanceimpacts,' and 62 percent expecting significantimpacts over the next 3 years. Although themethodology was not released publicly, it appearsthat the study did not include small firms, anddid not investigate the performance of non-TQMfirms over the same period.

The most widely-cited TQM research projectto date was the International Quality Study(American Quality Foundation, 1991), a jointproject conducted by Ernst & Young (theaccounting and consulting firm), and the Amer-ican Quality Foundation (the research armof ASQC, the American Society for Quality

20 T.C. Powell

Control). This project, which studied the TQMefforts of over 500 automotive, computer, bank-ing, and health care organizations in the U.S.,Canada, Germany, and Japan, had severalshortcomings: it had no theoretical grounding,the research sponsors had vested interests indisseminating TQM (their stated aim was 'todevelop an empirical basis for quality improve-ment worldwide'), it apparently excluded smalland medium-sized organizations, and it is notclear whether the study tracked the performanceof a control sample of non-TQM firms. Nonethe-less, by conducting in-depth field interviews andcontrolling for national and industry differences,this research was far more rigorous than itspredecessors. The sponsors concluded that someTQM practices—particularly process improve-ment and supplier certification—did universallyimprove performance, but the performanceimpacts of the remaining TQM features varieddepending on the firm's stage of TQM advance-ment. These results will be compared later withthe findings reported here.

Aside from these studies, which were mainlydesigned to show that TQM can work (with abit of consulting help), there exists a mutualfund—called the General Securities fund—thattrades only in stocks of firms known to adhereto the TQM philosophy. This fund has matchedthe performance of the Standard & Poor's 500despite a very conservative asset mix (the fundmanager keeps about 70% of fund assets incash), and Morningstar awarded a 4-star ratingto its 5-year performance. The stocks themselveshave slightly outperformed the S&P 500 over thesame period. Moreover, Business Week (1993)recently examined the stock performance of 10Baldrige winners, reporting that if a person hadinvested equal amounts in each Baldrige winnerwhen their awards were announced, the stockswould have appreciated a cumulative 89.2 percentsince 1988, compared to 33.1 percent for theStandard and Poor's 500. Although interesting,this evidence is inconclusive because it includesa small sample of very large firms, and it excludessmaller and privately-held firms (including thebankrupt Wallace Company). Moreover, sinceperformance is one criterion for Baldrige Awardselection (see Figure 1), the Business Weeksample was biased toward high-performing TQMfirms.

In sum, although several of the existing studies

claimed to prove that TQM does produceeconomic value, the question is not fully resolveddue to the methodological problems cited above,along with the contingencies raised in the Ernst& Young report, which is the most rigorousstudy to date.

TQM and imperfect imitability

Is TQM imitable? At first glance, it would appearso, and there is an impressive apparatus in placeto disseminate TQM to all U.S. organizations,including secondary schools and universities,health care institutions, financial services firms,law firms, CPA firms, nonprofits, and govern-ment. This apparatus includes the TQM gurus(with several best-sellers, 'Quality College' forexecutives, and regional training schools), thebusiness periodicals, consulting firms, best-selhngbooks and videos, executive education programs,college courses, the Baldrige Award publicity,and most importantly, the word-of-mouth testi-mony of executives already committed to thephilosophy. Many TQM firms ensure componentquality by requiring suppliers (including theirCPA firms and law firms) to adopt TQMprograms, and some (such as Ford, GM, andChrysler) have developed sophisticated suppliertraining and certification programs.

However, despite TQM's apparent widespreaddissemination—and the claims by adherents thatany firm can imitate TQM—there are powerfulreasons to believe TQM is imperfectly imitable.The diffusion of innovation literature provides auseful perspective on this issue. Whereas theresource literature focuses mainly on resourceimitation from the perspective of firms seekingto protect competitive advantage, the diffusionof innovation approach takes the perspective ofthe potential adopter. This line of research showsthat firms will not always attempt to imitateresources that produce advantages for competi-tors, and that diffusion of innovation depends onthe following factors (Rogers, 1983; Abrahamsonand Rosenkopf, 1993): (1) perceived relativeadvantage—the extent to which adopters believethe innovation is better than current practice;(2) compatibility—the degree to which an inno-vation is perceived by the adopter as consistentwith their needs, values, and experiences;(3) simplicity—the degree to which the innovationis perceived as understandable and


implementable; (4) trialability—the degree towhich an innovation can be experimented withon a limited basis; and (5) observability—thedegree to which an innovation and its benefitscan be observed by the potential adopter.

Diffusion of innovation models stress theimportance of similarity, or homophily, whichRogers (1983) defined as the degree to whichinnovator and potential adopter share attributessuch as objectives, strategies, norms, beliefs,experiences, and cultures. Empirical studies ina variety of disciplines have confirmed thatinnovations disseminate most rapidly amonghomophilous units, since they are most likelyto share perceptions of relative advantage,compatibility, simplicity, trialability, and observ-ability. It follows from the homophily principlethat heterophily (i.e., vital incongruities betweeninnovator and potential adopter) acts as a keybarrier to resource imitation.

Based on the five criteria for the adoption ofinnovations, one would not expect, a priori, thata high proportion of rationally-behaved U.S.firms would attempt to imitate TQM. Even TQMadvocates agree that TQM attributes confiictwith the existing philosophies and practices ofmost U.S. firms (e.g., Crosby, 1984). Theproblem intensifies among firms with feweropportunities to observe homophilous firms suc-cessfully engaged in R&D, such as small firmsand service businesses. Schaffer and Thomson(1992) have argued that six factors make TQMprograms especially difficult for U.S. firmsto imitate: (1) process (rather than results)orientation, (2) TQM is too large-scale anddiffused, (3) bad results are excused for the sakeof program success, (4) delusional measurementsof success, (5) TQM is staff- and consultant-driven, and (6) TQM is biased to orthodoxy, notcause and effect. The authors argued that, despitethe hype and hyperbole, TQM is ill-suited tomost U.S. firms, and that most should retaintheir traditional line-driven, results orientation.

Moreover, the Japanese origins of manyTQM practices produce an inherent heterophilybetween innovators and potential adopters thatdeter many U.S. firms. Young (1992) argues thatmanagement practices developed in Japan arefundamentally incompatible with U.S. managerialassumptions, observing that 'Japan has a uniquecultural and geographic history that affectsits institutions, assumptions about employee

behavior, business relationships, cost manage-ment, and performance evaluation systems. Thesesystems differ sharply from those in the UnitedStates' (1992: 678).

TQM appears to require sweeping reforms incore organizational features, particularly leader-ship styles and culture. Empirical studies conduc-ted from the organizational ecology perspectiveshow that innovations affecting core organiza-tional features such as strategy, structure, andculture pose the most significant survival risks,and may produce resistance to adoption even iftheir expected values are positive (Hannan andFreeman, 1984; Singh, Tucker, and House, 1986).As Carroll (1993) points out, many organizationswill resist attempting core reforms on the groundsof risk aversion, or on the expectation thatexisting strategies will presently bear fruit, orbecause they do not want to face the precarious-ness and turbulence of change, even if it is likelyto improve performance.

A resource-based analysis supports this con-clusion, suggesting that many potential adopterswould not find TQM readily imitable due totime compression diseconomies, connectednessof resources, causal ambiguity, and social com-plexity. Managers may applaud the notion ofemployee empowerment on the one hand, butfind quality action teams, suggestion systems,and training programs completely infeasiblewithin existing financial and human resources(e.g., long-term union members who regardexisting authority structures and work definitionsnot as stumbling blocks but as hard-earnedgains). In the study cited above. Young (1992)argued that firms are unlikely to adopt qualitypractices successfully in the short term if existingemployees lack work discipline, lack team orien-tation, lack cultural and demographic homogen-eity, prefer work rules, do not accept trainingwell, and are not accustomed to linkages betweencompensation and firm performance.

These characteristics not only take time tochange, but they highlight the difficulties raisedboth by complementary resources and causalambiguity. Potential TQM adopters may notappreciate that TQM success depends not onlyon adopting the TQM attributes, but also on thepreexistence of complementary factors apparentlyunrelated to TQM, yet more difficult to imitatethan TQM itself. For example, TQM appears torequire a culture receptive to change, a motivation

22 T.C. Powell

to improve, people capable of understanding andimplementing TQM's peculiar set of practices,corporate perseverance, leadership qualities suchas the capacity to commit, and perhaps someexogenous chance factor that may motivatechange and learning (e.g., threat from a foreigncompetitor). Without these tacit, intangible,causally-ambiguous, difficult-to-imitate com-plementary resources, TQM programs have nofoundation for success (Winter, 1987; Spender,1993; Szulanski, 1993).

It also appears that TQM requires a completerestructuring of social relationships both withinthe firm, and among the firm and its stakeholders.According to Barney (1991) 'A variety of firmresources can be socially complex. Examplesinclude the interpersonal relationships amongmanagers in a firm, a firm's culture, and afirm's reputation among suppliers and customers.'(1991: 110) Under TQM, firms must reconstituteall these relationships, in addition to relationshipsamong employees and between managers andemployees. And they must reconstitute themmore or less at the same time. As Barney pointsout, 'Such social engineering may be, for thetime being at least, beyond the capabilities ofmost firms. To the extent that such resourcesare not subject to direct management, theseresources are imperfectly imitable. (p. 110)

Despite widespread TQM adoption among theFortune 1000, and despite the bandwagon effects,publicity, and increasing external pressure, mostfirms have not adopted TQM. The one, highlypublicized group of U.S. firms that has adoptedTQM in overwhelming numbers—large manufac-turers competing in global markets—did so partlybecause of the powerful bandwagon (bolsteredby generally favorable performance studies), butprimarily because they, of all potential U.S.adopters, were most homophilous with the largeJapanese TQM innovators: they perceived TQMas providing advantage (as it had for the Japanesemanufacturers that threatened their markets),they believed it was compatible with existingneeds and experiences (the TQM gurus and earlyadopters asserted this, and Japanese firms likeToyota and Honda had shown TQM could workoutside Japan), they believed they had thenecessary complementary resources (e.g., a cul-ture amenable to employee empowerment), theybelieved they understood Japanese practices (theyused the same gurus, and some entered joint

ventures), and they perceived opportunities toexperiment with TQM on a trial basis (such asin GM's Cadillac Division). Qne would expecttheir success or failure with TQM to hinge almostentirely on the accuracy of these perceptions.

HYPOTHESES

If TQM does produce economic value, and thishas not been established, we would expect thepowerful isolating mechanisms just discussed toproduce competitive advantage for TQM firms.As such, we would expect to find that TQMfirms outperform non-TQM firms so long asfirms perceived their resources and capabilitieswith some reasonable degree of accuracy, andbehaved more or less rationally in decidingwhether to adopt TQM. Qf course, somefirms will act on wildly inaccurate perceptions(Starbuck, 1985)—in particular, some willwrongly believe they have the complementaryresources needed to make TQM work (such asthe capacity to commit). Qthers will behaveirrationally, joining the bandwagon withoutregard to TQM's demands or to their ownresources or suitability for TQM. Some firmsthat should adopt TQM will erroneously decideto reject it. Even if TQM does produce economicvalue, these inevitable mistakes may tend tosuppress any empirically-derived TQM-perform-ance relationship.

However, the diffusion of innovation literaturesuggests that, despite the inaccurate perceptionsand irrationalities that enter into all decisions(especially those involving bandwagonphenomena), most firms will assess themselvesand TQM more or less rationally using the fivecriteria, and those least-suited to TQM will notgenerally attempt it. As a result, if TQM doesproduce economic value, TQM firms shouldoutperform non-TQM firms overall. Therefore,the following hypothesis acts as an initial test ofthe economic value of TQM:

Hypothesis 1: TQM firms outperform non-TQM firms

A central notion in TQM training and literatureconcerns the need to adopt longer time horizons,and the expectation that TQM programs will notproduce short-term bottom-line results (Deming,


1986). Most TQM advocates agree that TQMcannot produce consistent performance advan-tages until after the third year of implementation(e.g., Schmidt and Finnigan, 1992), at whichpoint most organizations have had sufficient timeto adapt, assimilate, and stabilize under the newapproach. This notion suggests the followinghypothesis:

Hypothesis 2: Long-term TQM firms outper-form short-term TQM firms

The discussion of diffusion of innovations givesrise to the expectation that TQM would producethe best results for homophilous adopters, i.e.,those with experiences, conditions, and culturesresembling those of existing TQM practitioners.Since TQM originally emerged from manufac-turing environments, and because TQM remainsmore widely disseminated among manufac-turers, manufacturing adopters should outper-form service adopters. To some extent, such afinding could result from the time differencesjust discussed, since manufacturers have alonger history with TQM. The third hypothesispredicts that, independent of time effects,manufacturers outperform service firms amongTQM adopters.

Hypothesis 3: Manufacturing TQM firms out-perform service TQM firms

Finally, TQM performance should bear a positiveassociation with the firm's incorporation of the12 basic features of TQM. This capacity dependsboth on the firm's complementary resources atthe time of adoption, and on its ability to adaptand assimilate TQM principles over time. Thisleads to the remaining hypotheses:

Hypothesis 4: TQM performance is positivelyassociated with committed leadership

Hypothesis 5: TQM performance is positivelyassociated with adoption and communicationof TQM

Hypothesis 6: TQM performance is positivelyassociated with closer customer relationships

Hypothesis 7: TQM performance is positivelyassociated with closer supplier relationships

Hypothesis 8: TQM performance is positivelyassociated with benchmarking

Hypothesis 9: TQM performance is positivelyassociated with increased training

Hypothesis 10: TQM performance is posi-tively associated with open organization

Hypothesis 11: TQM performance is posi-tively associated with employee empowerment

Hypothesis 12: TQM performance is posi-tively associated with a zero-defects mentality

Hypothesis 13: TQM performance is posi-tively associated with flexible manufacturing

Hypothesis 14: TQM performance is posi-tively associated with process improvement

Hypothesis 15: TQM performance is posi-tively associated with measurement

DATA AND MEASURES

Sample

The empirical research proceeded in three phases.In the first phase, the researchers reviewedthe TQM literature, underwent TQM training,developed measurement scales for the TQMdimensions, and pretested these scales, includingreview and feedback from TQM consultants,trainers, and executives. In the second phase,the researchers mailed the pretested survey tothe CEQs of all firms with 50 employees ormore within selected zip codes in the northeasternU.S. The CEQs were asked to complete thesurvey whether or not their firms had adoptedTQM. In the third phase, the researchersconducted on-site personal interviews with CEQsand quality executives in 30 firms, also in selectedzip codes in the northeastern U.S., but not inzip codes included in the mail survey. Qf these30 firms, 23 had TQM programs, and these 23were also asked to complete the structuredsurvey.

The mail survey was designed and administeredunder guidelines established in Dillman's (1978)Total Design Method. Qf the 143 surveys mailedin the second phase, 40 were returned, 36 of

24 T.C. Powell

which were complete, for a usable response rateof 25.2 percent. Twenty-one of these firms(58.3%) claimed to have made meaningfulcommitments to TQM. This percentage is lowerthan those reported in earlier studies, butconsistent with the researchers' expectations ina representative sample including both manufac-turing and service firms. The researchers believethe percentage would have been much lower ifthe smallest firms had not been excluded fromthe sample.

In the third phase of the research, 19 of 23surveys were returned, 18 of which were complete(all from TQM firms, by design), for a usableresponse rate of 78.3 percent, and an overallusable response rate of 32.5 percent (54 responsesfrom 166 surveys). This response rate comparesfavorably wijh other studies using comparablemethodologies (e.g., Gomez-Mejia, 1992; Zahraand Covin, 1993).

Annual mediari sales among sample firmswas $136 million, and the median number ofemployees was 750. Although these statisticsexceeded known population parameters for allfirms located in those zip codes, this result hadbeen expected due to the artifical size minimumemployed in the sampling procedure. Thesemedians closely approximated those reported byPowell (1992) and Zahra and Covin (1993)using similar sampling procedures, and weresignificantly smaller than those found in someestablished data bases (e.g., the PIMS database). ' •

In the cover letter, the researchers requestedthat the survey be completed either by the CEQor, in the case of TQM firms, by either the CEQor a senior executive with overall responsibilityfor quality program implementation. Subsequentphone calls to a subsample of 10 phase tworespondents confirmed that the surveys werecompleted either by CEQs or senior qualityexecutives. All of the personal, on-site interviewsin the third phase included either CEQs, seniorquality executives, or both.

In the third phase, six firms were asked tocomplete two surveys per firm to establishinterrater reliability, and four firms responded.Among these respondents, the mean intrafirmcorrelation for the 92 survey items was 0.74,compared to 0.22 for interfirm responses. More-over, 76 percent of all intrafirm responses (210of 276) fell within a single point of one another

on the 5- and 6-point scales employed, comparedwith the 55 percent (152 of 276) that would beexpected by chance. Although the firm samplewas small, these statistics strongly supporteda presumption of interrater reliability in thesurveys.

Measures

Although TQM assessment instruments existedprior to this research (e.g., Saraph, Benson, andSchroeder, 1989; George, 1992), none was foundsuitable for this research, which required scalesthat integrated various approaches to TQM, ina form acceptable for scholarly survey researchand data analysis (for a critical evaluation ofexisting scales, see Human and Qhmer, 1993).In the pretest phase, the researchers developeda TQM measurement scale based on an exhaustivereview of the TQM literature, and revised thisscale through repeated discussions and site visitswith consultants and quality executives. The finalscale contained 47 items covering 12 variables,and is given in Appendix Ic.

Cronbach alpha coefficients were computed totest the reliabilities of the TQM scales (Cronbach,1951). Typically, these coefficients should fallwithin a range of 0.70 to 0.90 for narrowconstructs such as those defined here, and 0.55to 0.70 for moderately broad constructs (Van deVen and Ferry, 1979). In the empirical study,the coefficients for the twelve variables rangedbetween 0.78 and 0.90, and varied only triviallybetween the second and third phases of theresearch.

Resource-based strategy research has shownthat firm-level factors such as organization climate(Hansen and Wernerfelt, 1989) and structure(Powell, 1992) can explain performance varianceover and above variance attributable to industryfactors. To provide the most rigorous test of theperformance consequences of TQM, this researchassessed the TQM-performance relationship overand above industry and firm size effects. Tomeasure industry effects, the researchersdeveloped the scale of 14 industry items shownin Appendix la. Based on Porter's (1980) industryanalysis framework, these items were dividedinto two variables, entry barriers and rivalry,and these variables were used as an index ofindustry differences. Since each of these con-structs is somewhat broader than the TQM


constructs (e.g., rivalry includes advertising inten-sity, R&D, and industry growth rate), theCronbach coefficients of 0.64 and 0.60 wereconsidered acceptable under the Van de Venand Ferry (1979) criteria.

Qverall financial performance was measuredsubjectively, using the five items shown inAppendix Id, addressing profitability, salesgrowth, and overall financial performance. TQMperformance was assessed separately, using theitems shown in Appendix Ie. Subjective perform-ance measures are widely accepted in organiza-tional research (Lawrence and Lorsch, 1967;Dess, 1987; Powell, 1992), and in this researchwere preferred to financial statement data becausethe heterogeneous sample produced significantindustry differences in capital structures andaccounting conventions, and firm differences ininventory valuation, depreciation, and officers'salaries. Also, this research included manyprivately-held firms that would not have providedconfidential financial information as a matter ofpolicy.

As a test of the convergent validity of thetotal performance measure, objective financialmeasures were obtained for 15 publicly-heldsurvey participants. In this subsample, return onsales, a commonly-used measure of financialperformance in strategy research (e.g.. Cooland Dierickx, 1993; Zahra and Covin, 1993),correlated significantly with the subjectively-derived total performance measure (r = 0.64; p^ 0.01), suggesting that, although the objectiveand subjective measures are not identical, theobjective measures constituted a key element ofthe respondents' subjective assessments.

RESULTS

Table 3 shows the descriptive statistics andcorrelations among the variables involved intesting Hypothesis 1. The three TQM measuresare defined in Table 3, and the scales areprovided in Appendix 1. Table 3 indicates supportfor the validity of key measures used in thisresearch. For example. Table 3 shows that firmperformance correlated significantly with bothentry barriers (r = 0.29, p ^ 0.05) and rivalry(r = - 0.32; p < 0.05), and in the directionspredicted in the Porter framework. Moreover,these two variables explained independent pro-

portions of performance variance, correlatinginsignificantly with one another (r = - 0.09).Combined, the two industry variables explained17 percent of total performance variance, a resultconsistent with findings reported by Schmalansee(1985) and Rumelt (1991). Schmalansee foundthat 19.6 percent of profitability variance wasattributable to industry effects, concluding that(1985: 350) '80 percent of the variance inbusiness-unit profitability is unrelated to industryor share effects. While industry differencesmatter, they are clearly not all that matters.'Rumelt reported that 16.1 percent of profitabilityvariance was attributable to industry effects,although about half of this variance was unstablefrom one period to the next. Rumelt concludedthat (1991: 167) 'the most important sources ofeconomic rents are business-specific' The find-ings here support these conclusions.

Hypothesis 1 predicted that financial perform-ance would relate positively and significantlywith TQM. Table 3 shows that the three zero-order TQM-performance correlations were allpositive, and ranged from marginally significant(r = 0.23; p < 0.10) to highly significant (r =0.35; p s 0.01). However, the proper test ofthis hypothesis requires that industry and firmsize factors be partialled from the analysis. Table4 shows the results of this analysis, assessing thepartial correlations between TQM and perform-ance when industry and firm size have beenpartialled (see Cohen and Cohen, 1993).

Table 4 shows that all TQM-performancepartial correlations were positive and significant,ranging from pr = 0.30 (p < 0.05) to pr = 0.37(p < 0.01). Moreover, each partial correlationexceeded the corresponding zero-order corre-lation. This means that one or more of the threepartialled variables had suppressed the zero-order correlation through its joint correlationswith TQM and performance. An analysis of theTable 3 intercorrelations shows that the increasedTQM-performance partial correlations werecaused by entry barriers and firm size (In emps),and supports the following conclusions: (1) thezero-order TQM-performance correlations wouldhave been higher if TQM were not negativelycorrelated with entry barriers, which in turn waspositively correlated with performance. Althoughthe TQM-entry barrier correlation was notstatistically significant, and therefore does notrequire interpretation, the simple fact of its

26 T.C. Powell

Table 3. Descriptive statistics and correlations: Total sample

(N

1.2.3.4.5.6.7.

= 54)

Entry barriersRivalryIn EmpsTQMlTQM2TQM3Total performance

Mean

3.113.256.630.722.392.223.00

S.D.

0.860.522.220.451.751.521.15

1

1.00-0.09

0.03-0.11-0.10-0.04

0.29

2

1.000.01

-0.10-0.24-0.18-0.32

1,0.00

- 0

3

0020.21.10.13

4

1.000.850.920.23

5

1.000.910.28

6

1.000.35

7

1.00

TQM measures:TOMl—the dichotoitious variable shown in Appendix lb, item 1 (1 = TQM; 0 = no TOM).T0M2—the 6-point scale shown in Appendix lb, item 2.TQM3—mean response to all TOM items in Appendix lc except those for flexible mfg. (which do notapply to service firms).

negative direction was sufficient to create thesuppression effect; and (2) the zero-order TQM-performance correlations would have been higherif TQM were not positively (thoughinsignificantly) correlated with firm size, whichin turn correlated negatively with performance.

The findings in Table 4 support Hypothesis 1,along with the underlying assumption that TQMdoes provide economic value to the firm. If, asargued earlier, TQM is also difficult to imitatesuccessfully, then the results do not contradictthe assertion that TQM can act as a source ofcompetitive advantage, perhaps even sustainableadvantage. The remaining hypotheses considerthe conditions under which TQM may producesuch advantage.

Table 5 presents descriptive statistics andperformance correlations for the variablesinvolved in testing Hypotheses 12 through 15.Because these hypotheses concern only TQM

Table 4. Testing HI

{N = 54) Zero-order r Partial r

TQM variablesTQMlTQM2TQM3

0.23*0.28**0.35***

0.30**0.31**0.37***

*p < 0.10, '*p s 0.05, ***p < 0.01, tp £ 0.0011. All /-tests are two-tailed.2. Partial r is the correlation between Total Performance

and the TOM variable when the context set (entrybarriers, rivalry, and In emps.) is held constant.

adopters, only TQM firms {n = 39) were involvedin the remaining hypothesis testing. For eachindependent variable, correlations were com-puted for two performance measures, TQMPerformance (see Appendix Ie) and Total Per-formance (see Appendix Id). The former meas-ures satisfaction with the TQM program, andthe latter measures overall firm performance.One interesting but nonhypothesized result inTable 5 is that, even though large firms weremore likely to adopt TQM than small firms(see Table 3), the correlation between TQMperformance and firm size was significant andnegative, suggesting that size may impede success-ful TQM implementation.

Hypothesis 2 predicted that long-time TQMadopters would report better performance thanshort-time TQM adopters. The large zero-ordercorrelation between years since adoption andTQM performance (r = 0.53, p < 0.001) sup-ported this hypothesis, but the nonsignificantcorrelation with Total Performance (r = 0.10;ns) did not. This means that long-time TQMadopters were more satisfied with their TQMprograms than short-time adopters, even thoughno apparent time-performance correlationexisted. Tables 6 and 7 help to clarify thissituation, comparing the attributes of firms withmore than 4 years TQM experience to thosewith less than 4 years experience. These resultsshow that the long- and short-time adoptersdiffered significantly on six TQM variables, withtwo of them—the extent of training and processimprovement—very highly significant. A plaus-


Table 5. Statistics for testing H2-H15

39)

Contextual factorsH2:H3:

Years since TQM adoptionIndustry

In empsTQMH4:H5:H6:H7:H8:H9:HIO:Hll:H12:H13:H14:H15:

factorsExecutive commitmentAdopting the philosophyCloser to customersCloser to suppliersBenchmarkingTrainingOpen organizationEmp. empowermentZero defects mentalityFlexible mfg. (n = 24)Process improvementMeasurement

Mean

3.080.616.90

3.843.783.582.712.553.423.212.812.902.762.652.81

S.D.

2.010.492.09

0.861.040.881.120.851.000.960.991.290.890.931.13

Corr w/TQM Perf.

r

0.53t0.52t

-0.34**

0.41***0.29*0.27*0.61t0.240.43***0.61t0.64t0.61t0.53t0.57t0.56t

0.53t0.52t

-0.34**

0.36**0.170.250.34**0.020.210.61t0.59t0.45***

-0.010.210.18

Corr w/Total Perf.

r

0.100.04

-0.29*

0.45***0.32*0.30*0.28*0.220.200.51t0.45***0.34**0.010.220.19

Pr

0.100.04

-0.29*

0.36**0.26*0.200.30*0.220.160.43***0.37**0.28*0.040.250.17

*p s 0.10, **p a 0.05, ***p s 0.01, tp s 0.0011. All ^tests are two-tailed.2. pr is the correlation between performance and the variable given when the contextual factors are held constant. For thecontextual factors, r and pr are equivalent.3. Industry is a dichotomous variable (0 = service; 1 = mfg).

ible interpretation is that long-term TQM firmsreported more satisfaction with TQM becausethey had successfully mastered the core TQMtechniques, but they gained no significant per-formance advantages because they did notaccompany the techniques with improvements inthe intangibles that have significant performanceimpacts for all firms, particularly executivecommitment and open organizational culture.

Hypothesis 3 predicted that manufacturingTQM firms would outperform service TQMfirms. The large positive zero-order correlationbetween TQM performance and industry {r =0.52; p s 0.001) supported this hypothesis, butagain the nonsignificant partial correlation(r = 0.04) did not. The TQMperformance-industry correlation remained sig-nificant even after years since adoption had beenpartialled (r = 0.45; p s 0.001), suggesting thatmanufacturers were significantly more satisfiedwith their TQM programs than service firms,independent of any differences related to yearssince adoption. Tables 8 and 9 shed additionalhght on these findings, showing that manufac-turers differed significantly from service firms in

four areas—closeness to suppliers, zero defectsmentality, process improvement, andmeasurement—without differing in others thathave important performance impacts, such asopen organization, employee empowerment, andexecutive commitment. Again, it appears thatmanufacturers reported more satisfaction withtheir TQM programs because they successfullymastered the TQM tools and techniques, buttheir performance gains did not exceed those ofservice firms because the manufacturers did notsurpass them in the intangible areas mostresponsible for TQM performance.

Hypotheses 4 through 15 consider the relation-ships between performance and the 12 TQMvariables. These results are shown in the 'TQMfactors' section of Table 5, which shows both thezero-order correlations and the partial corre-lations when effects of the three contextualvariables have been removed. Qf the 12 TQMvariables, only three—executive commitment,open organization, and employeeempowerment—produced significant partial cor-relations for both performance measures. Twoadditional variables—zero-defects mentality and

28 T.C. Powell

Table 6. TQM performance and years since TQM adoption

Contextual factorsYears since TQM adoptionIndustryIn emps

TQM factorsExecutive commitmentAdopting the philosophyCloser to customersCloser to suppliersBenchmarkingTrainingQpen organizationEmp. empowermentZero defects mentalityFlexible mfg. (« = 24)Process improvementMeasurement

PerformanceTQM performance

Short-term (n

Mean

1.600.606.25

3.703.753.352.532.382.933.092.532.722.142.252.50

3.49

= 20)

S.D.

1.100.502.00

0.861.020.931.090.770.980.970.991.350.600.861.09

0.62

Long-term (n

Mean

4.840.687.37

4.043.973.842.972.834.003.403.203.233.283.143.25

4.07

= 19)

SD

1.010.481.92

0.841.030.771.070.900.790.910.781.180.760.751.02

0.65

A

tnsns

nsns*

ns

tns

ns

t

***

'p < 0.10, **p < 0.05, '**p s 0.01, tp < 0.0011. All /-tests are two-tailed.2. Short-term is defined as fewer than 4 years since adoption of TOM. Long-term is four years or more since adoption.

closeness to suppliers—correlated with TQMperformance atp s 0.05, but with total perform-ance only at p < 0.10. Therefore, of the Hypoth-eses 4 through 15, only Hypothesis 4 (executivecommitment). Hypothesis 10 (open organization),and Hypothesis 11 (employee empowerment)were supported conclusively. As noted earlier,these results suggest that the key to TQMperformance lies not in TQM tools and techniqueslike benchmarking and process improvement, butin intangible, behavioral factors like leadership,organizational skill, and culture.

To explore the possibility of TQM taxonomy,the researchers cluster-analyzed the 39 TQMfirms over the twelve TQM variables using aEuclidean, hierarchical, single-linkage clusteringalgorithm developed by Hartigan (1975). Thisprocedure produced the three groups shown inTable 10. Group 1 consisted principally of small-to medium-sized firms (median employees =500), mostly manufacturers (17 of 22 firms) withconsiderable TQM experience (median yearssince adoption = 4.0) and success (mean total

performance = 3.54). Group 2 consisted of 13firms, only two of which were manufacturers.These firms were quite large on average (medianemployees = 5,000), but had little experiencewith TQM (median years since adoption = 1.0years), and relatively poor performance (meantotal performance = 2.66). Group 3 consisted offour large manufacturing firms (medianemployees = 3,650) with considerable TQMexperience (median years since adoption = 3.5),but poor performance (mean total performance= 2.30).

Groups 1 and 3 each contain manufacturersexperienced in TQM, yet Group 1 is the highestperforming group, and Group 3 is the lowest.The mean differences in Table 10 suggest anexplanation. Although the Group 3 firms wereat least as advanced as Group 1 firms onseveral variables (process improvement, flexiblemanufacturing, closeness to suppliers, andbenchmarking), they fell behind even the TQM-inexperienced Group 2 firms in others (executivecommitment, adopting the philosophy, closeness


Table 7. TQM performance and years since TQM adoption

Correlations with TQM performance

Short-term (« = 20) Long-term (« = 19)

Contextual fact«-$IndustryIn emps

TQM factersExecutive commitmentAdopting the philosophyCloser to customersCloser to suppliersBenchmarkingTrainingQpen organizationEmp. empowermentZero defects mentalityFlexible mfg.Process improvementMeasurement

0.52**0.56***

0.150.080.020.44**0.030.350.53**0.49**0.45**0.040.190.43**

0.52**-0.56***

0.05-0.16-0.03

0.09-0.04

0.370.360.340.20

-0.03 (n =-0.01

0.26

11)

0.52**-0.41*

0.58***0.40*0.40*0.76t0.190.140.66***0.71t0.72t0.150.76t0.52**

0.52**-0.41*

0.280.180.310.69***0.150.110.49**0.59**0.53**0.39 (« =0.65***0.18

13)

*p £ 0.10, **p s 0.05. *'*p s 0.01. tp £ 0.0011. All Mests are two-tailed.2. r is the zero-order correlation between TQM Performance and the variable given.3. pr is the correlation between TOM Performance and the variable given when contextual factors are partialled. r = prlot contextual factors.4. Short-term is defined as fewer than 4 years since adoption of TQM. Long-term is 4 years or more since adoption.

to customers, open organization, and employeeempowerment). In other words, the Group 3firms mastered the imitable TQM tools—especially those directly related to production—without adopting its difficult-to-imitate intan-gibles. These programs more closely resembledtraditional Quality Control (QC) than contempo-rary TQM jwograms, and apparently had littlesignificant impact outside the factory.

DISCUSSION

The findings support the conclusion that TQMcan produce economic value to the firm, butthat it has not done so for all TQM adopters.TQM success appears to depend critically onexecutive commitment, open organization, andemployee empowerment, and less upon suchTQM staples as benchmarking, training, flexiblemanufacturing, process improvement, andimproved measurement. Although firms may find

these tools indispensible to a fully-integratedTQM initiative, they apparently do not produceperformance advantages in the absence of theintangibles. This result is consistent with theresource-based notion of complementaryresources, and suggests that, rather than merelyimitating TQM procedures, firms should focustheir efforts on creating a culture within whichthese procedures can thrive. This profile differssubstantially from those found in existing TQMstudies, which focused almost entirely on theTQM tools and techniques.

Moreover, TQM .does not appear to be aspervasive as some earlier studies suggested: alittle over half of firms responding to the phase-2 survey had adopted TQM, and this numberwas overstated due to the minimum firm size (50employees) imposed in the sampling design.Although this result differed from the 93 percentand 74.5 percent adoption rates reported amongthe largest U.S. firms, it was in keeping with theresearchers' expectations of potential adopters

30 T.C. Powell

Table 8. TQM performance and industry group

Contextual factorsYears since TQM adoptionIndustryIn emps

TQM factorsExecutive commitmentAdopting the philosophyCloser to customersCloser to suppliersBenchmarkingTrainingQpen organizationEmp. empowermentZero defects mentalityFlexible mfg. (n = 24)Process improvementMeasurement

PerformanceTQM Performance

Mfg. (n

Mean

3.541.006.64

3.944.003.533.282.683.533.382.983.352.763.033.35

4.05

= 24)

S.D.

1.720.002.03

0.900.910.940.860.840.870.910.971.000.890.800.86

0.62

Service (n

Mean

2.330.007.37

3.673.433.651.802.333.232.932.552.18na

2.041.95

3.32

= 15)

S.D.

2.260.001.92

0.781.180.800.870.851.191.020.991.41na

0.790.98

0.56

A

nstns

ns*

nstnsnsnsns***natt

t

•p s 0.10, *'p s 0.05, ***pAll Mests are two-tailed.

0.01, tp < 0.001

who evaluated TQM more or less rationally inlight of their own needs and resources.

The results suggest that, although the intan-gibles were universally important to TQM success,other factors were context-dependent. In parti-cular, closer supplier relationships appeared topromote TQM performance among manufac-turers but not among service firms, and processimprovement appeared to promote TQM per-formance among service firms but not amongmanufacturers. These results support the contin-gency approach adopted in the American QualityFoundation (1991) study, but they tell a slightlydifferent story. The earlier study reported thatsupplier certification and process improvementwere uniformly associated with performanceacross all respondent groups. The results reportedhere suggest that the performance-supplierrelationship, though significant, vanishes whenindustry and years since adoption are partialled(see Table 5), although it remains significant forthe subsample of manufacturers (see Table 9).Similarly, the performance-process improvementcorrelation, though significant, vanishes when

industry and years since adoption are partialled,although it remains significant for service firms.In other words, when industry and time aretaken into account, closeness to suppliers andprocess improvement do not generally explainTQM performance variance, although they mayexplain performance in some industry contexts.

The on-site interviews conducted in the thirdphase of this research produced anecdotal evi-dence that supplements this profile of TQM asa strategic resource. During this phase of thestudy, the researchers encountered several highlysuccessful non-TQM firms. Qne firm that typifiedthis group had produced remarkable growthand profitability in a rivalrous, low technologymanufacturing industry, although it explicitlyrejected TQM. The interviews showed that thisfirm had for years done many of the thingspromoted by TQM advocates—includingempowering employees, eliminating bureaucracy,and simplifying processes—but had neverregarded these activities as an integrated formalprogram to be adopted, and had not applied aspecial vocabulary or ideology to them. These


Table 9. TQM performance and industry group

Correlations with TQM performance

Mfg. {n = 24) Service (n = 15)

Contextual factorsYears since TQM adoptionIn emps

TQM factorsExecutive commitmentAdopting the philosophyCloser to customersCloser to suppliersBenchmarkingTrainingQpen organizationEmp. empowermentZero defects mentalityFlexible mfg.Process improvementMeasurement

0.51***0.26

0.55***0.220.48**0.53***0.100.42**0.57***0.69t0.53***0.330.300.43**

0.51***-0.26

0.320.140.180.59***0.110.090.47**0.57***0.310.290.150.27

0.43*-0.37

0.060.140.120.230.250.45*0.63t0.54***0.47*

na0.60***0.22

0.43-0.37

-0.150.02

-0.15-0.05

0.020.340.56***0.280.37

na0.54*0.00

•p s 0.10, **p s 0.05, **'p < 0.01, tp s 0.0011. All Mests are two-tailed.2. r is the zero-order correlation between TQM performance and the variable given.3. pr is the correlation between TQM performance and the variable given when contextual factors are partialled. r = pr

for contextual factors.

activities were, in the words of its CEQ, just'common sense' and 'good business.'

The empirical results suggested that TQM canproduce competitive advantage, but is TQMnecessary to success? Apparently not. Both theanecdotal and statistical evidence suggest that,although TQM can produce competitive advan-tage, adopting the vocabularies, ideologies, andtools promoted by the TQM gurus and advocatesmatters less than developing the underlyingintangible resources that make TQM implemen-tation successful. And these resources appear toproduce success with or without formal TQMadoption: TQM firms that lack them do notsucceed, and non-TQM firms that have them do.Perhaps TQM's highest purpose, and its realcontribution to American business, is in providinga framework that helps firms understand andacquire these resources as part of an integratedchange program. Qne executive in a successfulnon-TQM firm put it best: 'If a company needsa fancy program to listen to their customers,then I think they'd better get one.

CONCLUSIONS

In the empirical study, the researchers employeda variety of validating procedures and controls,including extensive pretesting, confirming theidentities of a respondent sample, testing forinterrater relaibility, testing the convergent va-lidity of the total performance measure, con-ducting on-site personal interviews, comparingsample statistics with population parameters andwith statistics reported in comparable studies,testing all measurement scales for internalreliability, focusing on a single geographic area(controlling for regional performancedifferences), and controlling for factors that mayproduce spurious zero-order correlations (suchas industry and firm size). Nonetheless, thesurvey methodology has several limitations thatshould be addressed in interpreting the findings.

Qne limitation is the study's cross-sectionalresearch design. Although the data showed asignificant TQM-performance correlation, theydid not strictly prove that TQM caused perform-

32 T.C. Powell

Table 10. Clustering firms on TQM variables

Contextual factorsMedian years since TQMIndustryMedian emps

TQM factorsExecutive commitmentAdopting the philosophyCloser to customersCloser to suppliersBenchmarkingTrainingQpen organizationEmp. empowermentZero defects mentalityFlexible mfg.Process improvementMeasurement

PerformanceTotal performanceTQM performance

Cluster 1

(n = 21)Mean

4.000.81

500.0

4.334.433.983.272.783.953.743.433.652.543.103.54

3.544.11

Cluster 2

(« = 13)Mean

1.000.23

5000.0

3.182.963.121.741.952.652.562.001.490.711.791.77

2.663.29

Cluster 3

{n = 5)Mean

3.501.00

3650.0

3.082.752.633.253.173.062.061.813.003.003.402.81

2.303.56

Multiple

1-2

***

t**

tt

***t

***ttttttt

**t

comparisons

1-3

nsns***

***t

***nsns**

tt**nsnsns

***

2-3

nsnsns

nsnsns******nsnsnsttt

***

nsns

*p < 0.10. **p s 0.05. 'All f-tests are two-tailed.

p < 0.01. tp £ 0.001

ance to increase, but only that an associationexisted. High performance may give rise to TQMprograms, or TQM and performance may bothbe caused by some third factor not measured inthis study (although, based on previous strategyresearch, the most powerful known explanatoryfactors were included in the study). Theresearchers believe that causation from TQM toperformance is the most plausible of the variousinterpretations, but a longitudinal design—withpre- and post-TQM performance measures—would be required to support a causal inference.

Another limitation is the relatively small samplesize employed in the study. The samples of allfirms {n = 54) and all TQM firms (n = 39)were quite sufficient for the statistical methodsemployed in this study, and the sample statisticsstrongly supported the integrity of the sample, asdid other sample demographics, which conformedboth with anecdotal evidence and with findingsin other studies (e.g., TQM firms were, onaverage, larger than non-TQM firms, and service

firms trailed manufacturing firms in TQMimplementation). However, the subsamples ofmanufacturing TQM firms (n = 24), service TQMfirms (n = 15) and non-TQM firms (n = 15) werenot large, and the findings from the associatedhypothesis tests should be generalized withcaution. The researchers acknowledge the poten-tial external validity problems in the subsamples,and urge other strategy researchers to test theirfindings using larger samples and alternativemethodologies.

Qne concern of the researchers was that,although the survey cover letter asked all firmsto respond, whether or not they had adoptedTQM, the subject matter of the study (TQM)may have produced greater nonresponse amongnon-TQM firms than among TQM firms. Asnoted earlier, 58 percent of the phase two mailsurvey respondents were TQM firms, and 42percent were non-TQM. Although this indicateslower TQM dissemination than previous studies,the researchers believe, based mainly on anec-


dotal and on-site interview evidence, that theactual proportion of TQM firms in the populationis much lower than 58 percent, and perhapscloser to 30 percent. The 58 percent proportionin the sample was in part due to the sizeminimum of 50 employees imposed on thesample, and perhaps partly due to higher responseamong TQM firms. In any case, there is noreason to believe this factor would invalidate anyof the empirical results reported here (other thanthe 58%-42% split).

A final limitation was survivor bias, althoughthe researchers did attempt to address this in theresearch design. From an organizational ecologyperspective, one could argue that the studyunderstates the TQM-performance relatiolishipbecause non-TQM firms are failing at a fasterrate than TQM firms. If this is true, then TQMwill continue to disseminate rapidly not onlybecause more firms are adopting TQM, butbecause the failure of non-TQM firms will leavean ever-rising proportion of TQM firms in thesurviving population. This research did notstudy nonsurvivors, and the researchers did notencounter studies comparing survival rates amongTQM and non-TQM firms. As noted earlier, theresearchers have concluded that survival isassociated with certain critical intangibleresources, whether or not they fall within a TQMinitiative.

The researchers believe this study containsfindings useful both to practicing managers andto other strategy researchers. The message formanagers is that, although TQM programs canproduce performance advantages, they do notaddress the needs of all organizations, and theyare fraught with pitfalls for firms that lack therequisite complementary resources. Moreover,despite relentless pressure from TQM advocates,we have concluded that it is quite possible forfirms to prosper outside the confines of the TQMideology and vocabulary, so long as they nurturethe intangible resources critical to survival andsuccess.

This study makes several contributions tostrategic management research. First, in contrastto the existing TQM literature, which is almostentirely practitioner-oriented, this study providesthe first integrated theoretical and empiricalprofile of TQM as a strategic resource. Theresearchers know of no other managementconcept or practice that has ever received

so much practitioner attention, with so littleacademic study, as TQM, and this study beginsto redress that imbalance. Second, this studyprovides support for the expanding stream ofresource-based empirical research that demon-strates the importance of firm-level intangibleresources. Although these resources pose difficultmethodological and measurement problems, thisline of research is now showing signs of maturityand accumulation. Third, this study integratesresource-based insights with those from alterna-tive theoretical perspectives, specifically diffusionof innovation theory and organizational ecology.Although the notion of a unified, resource-basedstrategic management paradigm has appeal, thisstudy shows that multiple theoretical approachescan still add depth to the interpretation ofempirical data.

ACKNOWLEDGEMENTS

This research could not have been completedwithout the enthusiastic and capable researchassistance of Alissa Robinson. The author wouldalso like to acknowledge financial support pro-vided by Bryant College.

REFERENCES

Abrahmson, E. and L. Rosenkopf (1993). 'Institutionaland competitive bandwagons: Using mathematicalmodeling as a tool to explore innovation diffusion'.Academy of Management Review, 18(3),pp. 487-517.

Akao, Y. (1991). Hoshin Kanri: Policy Deploymentfor Successful TQM. Productivity Press, Cambridge,MA.

American Quality Foundation and Ernst & Young(1991). International Quality Study: The DefinitiveStudy of the Best International Quality ManagementPractices. Ernst & Young, Cleveland, OH.

Arthur D. Little Corporation (1992). 'Executivecaravan TQM survey summary', private correspon-dence dated Qctober 15, 1992.

Barney, J. (1986a). 'Qrganizational culture: Can it bea source of sustained competitive advantage?'Academy of Management Review, 11(3),pp. 656-665.

Barney, J. (1986b). 'Strategic factor markets: Expec-tations, luck, and business strategy,' ManagementScience, 32 (10) pp. 1231-1241.

Barney, J. (1991). 'Firm resources and sustainedcompetitive advantagee', Journal of Management,17, pp. 99-120.

34 T.C. Powell

Benson, T. (July 5, 1993). 'Quality: If at first youdon't succeed'. Industry Week, pp. 48-59.

Bleakley, F. (July 6, 1993). 'Many companies trymanagement fads, only to see them flop'. WallStreet Journal, p. Al.

Business Week (October 18, 1993). 'Betting to win onthe Baldie winners', p. 8.

Carroll, G. (1993). 'A sociological view on whyfirms differ'. Strategic Management Journal, 14(4),pp. 237-249.

Cohen, J. and P. Cohen (1983). Applied MultipleRegression-Correlation Analysis for the BehavioralSciences. Erlbaum, Hillsdale, NJ.

Cool, K. and I. Dierickx (1993). 'Rivalry, strategicgroups, and firm profitability'. Strategic ManagementJournal, 14(1), pp. 47-59.

Cool, K. and D. Schendel (1988). 'Performancedifferences among strategic group members'. Stra-tegic Management Journal, 9(3), pp. 207-223.

Cronbach, L. (1951). 'Coefficient alpha and theinternal structure of tests', Psychometrica, 16,pp. 297-334.

Crosby, P. (1979). Quality is Free: The Art of MakingQuality Certain. McGraw-Hill, New York.

Crosby, P. (1984). Quality without Tears: The Art ofHassle-Free Management. McGraw-Hill, New York.

Deming, W. E. (1986). Out of the Crisis. MIT Centerfor Advanced Engineering Study, Cambridge, MA.

Dess, G. (1987). 'Consensus on strategy formulationand organizational performance: Competitors in afragmented industry'. Strategic Management Jour-nal, 8(3), pp. 259-277.

Dierickx, I. and K. Cool (1989). 'Asset Stockaccumulation and sustainability of competitiveadvantage'. Management Science, 35(12),pp. 1504-1513.

Dillman, D. (1978). Mail and Telephone Surveys: TheTotal Design Method. Wiley, New York.

Fiol, M. (1991). 'Managing culture as a competitiveresource: An identity-based view of sustainablecompetitive advantage'. Journal of Management,17, pp. 803-813.

Fuchsberg, G. (May 14, 1992). 'Quality programsshow shoddy results'. Wall Street Journal, p. Bl.

Fuchsberg, G. (April 19, 1993a). 'Baldrige award maybe losing some luster'. Wall Street Journal, p. Bl.

Fuchsberg, G. (August 26, 1993b). 'Small firmsstruggle with latest management trends'. Wall StreetJournal, p. B2.

Garvin, D. (July-August, 1993). 'Building a learningorganization'. Harvard Business Review, pp. 78-91.

George, S. (1992). The Baldrige Quality System.Wiley, New York.

Gomez-Mejia, L. (1992). 'Structure and process ofdiversification, compensation strategy, and firmperformance. Strategic Management Journal, 13(5),pp. 381-397.

Grayson, C. J. and C. Q'Dell (1988). AmericanBusiness: A Two-Minute Warning. Free Press, NewYork.

Hall, R. (1992). 'The strategic analysis of intangibleresources'. Strategic Management Journal, 13(2),pp. 135-144.

Hannan, M. and J. Freeman (1984). 'Structural inertiaand organizational change', American SociologicalReview, 49, pp. 149-164.

Hansen, G. and B. Wernerfelt (1989). 'Determinantsof firm performance: The relative performance ofeconomic and organizational factors'. StrategicManagement Journal, 10(5), pp. 399-411.

Hartigan, J. (1975). Clustering Algorithms. John Wiley& Sons, New York.

Hayes, R. and W. Abernathy (July-August 1980).'Managing our way to economic decline'. HarvardBusiness Review, pp. (H-ll.

Hill, R. (1993). 'When the going gets tough: ABaldrige Award winner on the line'. The Executive,7(3), pp. 75—79.

Human, S. and D. Qhmer (1993). 'Desperately seekingstrategy: An analysis of total quality managementassessment instruments using Porter's (1985) valuechain'. University of Kentucky working paper.

Imai, M. (1986). Kaizen: The Key to Japan's Competi-tive Success. McGraw-Hill, New York.

Ishikawa, K. (1985). What is Total Quality Control?The Japanese Way. Prentice-Hall, Englewood Cliffs,NJ.

Jacobs, M. (1991). Short-Term America. HarvardPress, Cambridge, MA.

Juran, J. (1988). Juran on Planning for Quality.American Society for Quality Control, Milwaukee,WI.

Juran, J. (1992). Juran on Quality by Design. FreePress, New York.

Juran, J. (July-August, 1993). 'Made in U.S.A.: Arenaissance in quality'. Harvard Business Review,pp. 42-50.

Lawless, M., D. Bergh, and W. Wilsted (1989).'Performance variations among strategic groupmembers: An examination of individual firm capa-bility', youma/o/Manage/nenr, 15(4), pp. 649-651.

Lawrence, P. and J. Lorsch (1967). Organizationand Environment. Harvard Business School Press,Boston, MA.

Lieberman, M. and D. Montgomery (1988). 'Firstmover advantages'. Strategic Management Journal,Summer Special Issue, 9, pp. 41-58.

Mathews, J. (September 7, 1992). 'The cost of quality',Newsweek, pp. 48-49.

Naj, A. (May 7, 1993). 'Some manufacturers dropefforts to adopt Japanese manufacturing tech-niques'. Wall Street Journal, p. Al.

Pascale, R. (1981). The Art of Japanese Management.Simon & Schuster, New York.

Peteraf, M. (1993). The cornerstones of competitiveadvantage: A resource-based view'. Strategic Man-agement Journal, 14(3). pp. 179^191.

Porter, M. (1980). Competitive Strategy: Techniquesfor Analyzing Industries and Competitors. FreePress, New York.

Powell, T. (1992). 'Organizational alignment as com-petitive advantage'. Strategic Management Journal,13(2), pp. 119-134.

Rogers, E. (1983). Diffusion of Innovations. FreePress, New York.

Ross, J. (1993). Total Quality Management: Text,


Cases and Readings. St. Lucie Press, Delray Beach,FL.

Rumelt, R. (1984). 'Toward a strategic theory of thefirm'. In R. Lamb (ed.). Competitive StrategicManagement, Prentice-Hall, Englewood Cliffs, NJ,pp. 556-570.

Rumelt, R. (1991). 'How much does industry matter?'.Strategic Management Journal, 12(3), pp. 167-185.

Saraph, J., G. Benson and R. Schroeder (1989). 'Aninstrument for measuring the critical factors ofquality management'. Decision Sciences, 20,pp. 810-829.

Schmalansee, R. (1985). 'Do markets differ much?',American Economic Review, 75(3), pp. 341-351.

Schaffer, R. and H. Thomson (January-February,1992). 'Successful change programs begin withresults'. Harvard Business Review, pp. 80-89.

Schmidt, W. and J. Finnigan (1992). The Race withouta Finish Line: America's Quest for Total Quality.Jossey-Bass, San Francisco, CA.

Senge, P. (1990). The Fifth Discipline. Doubleday,New York.

Singh, J., D. Tucker, and R. House (1986). 'Organiza-tional change and organizational mortality'. Admin-istrative Science Quarterly, 31, pp. 587-611.

Spechler, J. (1991). When America Does it Right.Industrial Engineering and Management Press,Norcross, GA.

Spender, J. (1993). 'Competitive advantage from tacitknowledge? Unpacking the concept and its strategiclimitations'. Academy of Management Best PaperProceedings, pp. 37-41.

Stalk, G., P. Evans and L. Shulman (March-April1992). 'Competing on capabilities: The new rulesof corporate strategy'. Harvard Business Review,pp. 57-69.

Stalk, G. and T. Hout (1990). Competing AgainstTime: How Time-Based Competition Is ReshapingGlobal Markets. Free Press, New York.

Starbuck, W. (1985). 'Acting first and thinking later:Finding decisions and strategies in the past'. In J.Pennings and Associates, Organizational Strategyand Change. Jossey-Bass, San Francisco, CA,pp. 336-372.

Szulanski, G. (1993). 'Intra-firm transfer of bestpractice, appropriable capabilities, and organiza-tional barriers to appropriation'. Academy ofManagement Best Paper Proceedings, pp. 47-51.

U.S. General Accounting Qffice (1991). ManagementPractices: U.S. Companies Improve Performancethrough Quality Efforts. U.S. General AccountingQffice, Gaithersburg, MD.

Van de Ven, A. and D. Ferry (1979). Measuring andAssessing Organizations. Wiley, New York.

Walton, M. (1986). The Deming Management Method,Pedigree, New York.

Weigelt, K. and C. Camerer (1988). 'Reputation andcorporate strategy: A review of recent theory andapplications'. Strategic Management Journal, 9(5),pp. 443-454.

Wernerfelt, B. (1984). 'A resource-based view ofthe firm'. Strategic Management Journal, 5(2),pp. 171-180.

Winter, S. (1987). 'Knowledge and competence asstrategic assets'. In D. Teece (ed.). The CompetitiveChallenge. Center for Research in Management.Berkeley, CA, pp. 159-184.

Young, M. (1992). 'A framework for the successfuladoption of Japanese manufacturing techniques inthe United States', Academy of ManagementReview, 17(4), pp. 677-700.

Zahra, S. and J. Covin (1993). 'Business strategy,technology policy, and firm performance'. StrategicManagement Journal, 14(6), pp. 451-478.

APPENDIX 1: MEASUREMENT SCALES

la: INDUSTRY

Participants were asked to circle the best responseto each statement on a 1-5 scale (5 = agreestrongly, 1 = disagree strongly). The variablesmeasured are indicated for all scales, but werenot so indicated in the surveys.

Entry barriers

1. Our industry is very difficult for new firms toenter successfully

2. In our business, existing firms have insur-mountable advantages over new entrants

3. Large firms have definite cost advantages inour industry

4. Our industry is dominated by a few largecompetitors

Rivalry

1. In our industry, customers are loyal—theyrarely switch to new firms or competitors

2. Competition in our industry is mainly onprice, not product or service differentiation

3. Compared to other industries, rivalry in ourindustry is extremely intense

4. Firms in our industry advertise heavilycompared to other industries

5. Demand in our industry has been growingrapidly in the past 3 years

6. Innovation and R&D are more prevalent inour industry than in most industries

7. Over the past 3 years, our industry has beenmore profitable than most

8. We have a serious excess capacity problemin our industry

9. Our industry is still in early growth andinfancy

36 T.C. Powell

10. Our industry would be characterized as ahigh-technology industry

3. Executives actively communicating a Quahtycommitment to employees

lb: TQM GENERAL

Respondents were asked the following twogeneral TQM questions:1. Has your organization ever made a significant

commitment to Total Quality Management ora similar Total Quality program?

A. YESB. NQ2. Please tell us how advanced the implemen-

tation of the program is compared to Qualityprograms of other organizations you arefamiliar with.

A. FAR MQRE ADVANCED in implemen-tation than most other organizations I amfamiliar with

B. SQMEWHAT MQRE ADVANCED inimplementation than most other organizationsI am familiar with

C. ABQUT EQUALLY ADVANCED inimplementation than most other organizationsI am familiar with

D. SOMEWHAT LESS ADVANCED inimplementation than most other organizationsI am familiar with

E. FAR LESS ADVANCED in implementationthan most other organizations I am familiarwith

F. NQ SIGNIFICANT INVQLVEMENT witha Quality program

lc: TQM FACTORS

Respondents were asked to indicate theirimplementation of the Quality features givenbelow, on a 0 to 5 scale (5 = highly advancedin implementation; 1 = have not begun implemen-tation but intend to; 0 = do not intend toimplement).

Adopting the philosophy

1. Quality principles included in our missionstatement

2. An overall theme based on our Qualityprogram

3. Entering a Baldrige Award competition

Closer to customers

L Increasing the organization's direct personalcontacts with customers

2. Actively seeking customer inputs to determinetheir requirements

3. Using customer requirements as the basis forQuality

4. Involving customers in product or servicedesign

Closer to suppliers

1. Working more closely with suppliers2. Requiring suppliers to meet stricter Quality

specifications3. Requiring suppliers to adopt a Quality pro-

gram

Benchmarking

1. An active competitive benchmarking program2. Researching best practices of other organiza-

tions3. Visiting other organizations to investigate best

practices first hand

Training

1. Management training in Quality principles2. Employee training in Quality principles3. Employee training in problem-solving skills4. Employee training in teamwork

Executive commitment

L A top executive decision to commit fully to aOuality program

2. Top executives actively championing ourQuality program

Open organization

L A more open, trusting organizational culture2. Less bureaucracy3. Frequent use of cross-departmental teams4. Use of empowered work teams


Employee empowerment

1. Increased employee involvement in designand planning

2. A more active employee suggestion system3. Increased employee autonomy in decision-

making4. Increased employee interaction with customers

and suppliers

Zero-defects mentality

1. An announced goal of zero-defects2. A program for continuous reduction in defects3. A plan to reduce rework drastically

Flexible manufacturing

1. Design for Assembly (DFA) or Design forManufacturability (DMA)

2. A flexible manufacturing system3. A just-in-time inventory system4. Cellular manufacturing5. Process capability studies6. Statistical Process Control7. Taguchi methods, or Design of Experiments

(DQE)

Process improvement

1. A program to reduce order-processing cycletime

2. A program to reduce new product or servicedevelopment cycle times

3. A program to reduce overall product orservice delivery cycle times

4. A program to reduce paperwork5. A program to find wasted time and costs in

all internal processes

Measurement

1. Measurement of Quality performance in allareas

2. Charts and graphs to measure and monitorQuality

3. Statistical methods to measure and monitorQuality

4. Employee training in statistical methods formeasuring Quality

Id: TOTAL PERFORMANCE

Using the agree-disagree scale shown earlier,respondents were asked to rate their firms'overall performance over the past 3 years:

1. Qver the past 3 years, our financial perform-ance has been outstanding

2. Qver the past 3 years, our financial perform-ance has exceeded our competitors'

3. Qver the past 3 years, our revenue (sales)growth has been outstanding

4. Qver the past 3 years, we have been moreprofitable than our competitors

5. Qver the past 3 years, our revenue growthrate has exceeded our competitors'

Ie: TQM PROGRAM PERFORMANCE

Using the agree-disagree scale shown earlier,respondents were asked to indicate how theirTQM programs had impacted performance byanswering each question.

1. Qur Quality program has dramaticallyincreased our organization's productivity

2. Qur Quality program has improved ourcompetitive position

3. Qur Quality program has dramaticallyincreased our profitability

4. Qur Quality program has dramaticallyincreased our revenues

5. Qur Quality program has dramaticallyimproved our overall performance

6. Qur Quality program has been a positivedevelopment for our organization

7. Qur Quality program has had a negativeimpact on our profitability

8. We would have been better off without aQuality program

total quality management as competitive advantage

Documents