1、The Rhetoric and Realityof Total QualityManagementMark J. ZbarackiUniversity of Chicago by Cornell University,0001-8392/98/4303-0602/$! ,00,I thank Kimberly D, Elsbach, James V,Jucker, James G, March, Gerardo Okhuy-sen, and Robert I, Sutton for their helpon previous versions of this paper, i alsotha
2、nk the Stanford Center for Organiza-tional Research, Stanford Universitys Of-fice of Technology Licensing Researchincentive Fund, the Stanford IntegratedManufacturing Association, and the Uni-versity of Chicago Graduate School ofBusiness for supporting this research.Special thanks to Mark Bergen, St
3、ephenSchioesser, and George Wu, as well asChristine Oliver and the three anonymousreviewers for very helpful advice in devel-oping the ideas in this version. I pre-sented an earlier version of this-paper inApril 1994 at the Society for Industrialand Organizational Psychology in Nash-viRe, Tennessee
4、and in August 1994 atthe annual meetings of the Academy ofManagement in Dallas, Texas,This article induces a model of the evolving rhetoric andreality of total quality management (TQM) in five organi-zations to show how institutional forces can distort thetechnical reality of TQM. Using interviews,
5、organizationaldocuments, and observation, t follow the social construc-tion of TQM in these organizations to trace the relation-ship between the technical practices and rhetoric ofTQM. The model shows that managers consume a rheto-ric of success about TQM, use that rhetoric to developtheir TQM progr
6、am, and then filter their experiences topresent their own rhetoric of success. Consequently, thediscourse on TQM develops an overly optimistic view ofTQM. The models demonstrate how individual actionsand discourse shape TQM and fuel institutional forces.*We are absolutely convinced that TQM is a fun
7、damentally bet-ter way to conduct business and is necessary for the economicwelt-being of America. TQM results in higher-quality, lower costproducts and services that respond faster to the needs of thecustomer.The CEOs of American Express Company, IBM Corporation,The Procter and Gamble Company, Ford
8、 Motor Company, Mo-torola Inc., and Xerox Corporation in an open letter sent to Har-vard Business Review in 1991The term is counterproductive. My work is about a transforma-tion in management and about the profound knowledge neededfor the transformation. Total quality stops people from thinking.W. E
9、dwards Deming (quoted in Senge, 1992)The phenomenal spread of total quality nnanagement (TQM)has generated an ironic controversy. The controversy pitsTQM advocates, who see it as a uniquely effective methodfor improving organizational performance, against opponents,who see it only as the latest of m
10、any organizational fads(Hackman and Wageman, 1995). The irony is that the contro-versy sets advocates of TQM against scholars whose exper-tise encompasses the very roots of the TQM methods. Ad-vocates of TQM have used TQM to build a position fromwhich they criticize academics for their failure to st
11、udy TQM(Robinson et al., 1991). Yet those academics should under-stand TQM better than the advocates, because TQM em-ploys technical methods scholars have studied for years(Dean and Bowen, 1994; Hackman and Wageman, 1995).Meanwhile, some of the original experts in the quality move-ment, who continue
12、 to preach a renewal of business, havecome to detest the term total quality management (Senge,1992). And as organizational scholars have scrambled to un-derstand the TQM phenomenon, they find themselves grap-pling with diffuse and ambiguous definitions of TQM. Mostorganizational scholars who have re
13、sponded to the call forTQM research have focused their theoretical efforts on refin-ing definitions of TQM (Dean and Bowen, 1994; Sitkin, Sut-cliffe, and Schroeder, 1994; Spencer, 1994; Hackman andWageman, 1995) and of quality (Reeves and Bednar, 1994).The controversy has created two competing rheto
14、rical posi-tions and great difficulty reconciling the two.Such problems are not unique to TQM. Empirically, variousmanagement fads suggest that TQM represents the latestinstance of an enduring problem. Hackman (1975), near thepeak of his work on job enrichment, predicted its comingdemise. Lawler and
15、 Mohrman (1985) triggered an explosion602/Administrative Science Quarterly, 43 (1998): 602-636Rhetoric uid Realityof controversy by anticipating the decline of quality circles.Zipkin 1991 puzzled over the overiy enthusiastic and inap-propriate use of just-in-time (JIT) manufacturing. And, morerecent
16、ly, reengineering seems to have followed the samepattern of expansive rhetoric unmet by the reality of use.Other examples abound, including T-groups and manage-ment-by-objectives (Hackman and Wageman, 1995). Theoret-ically, recent research on management fashion Abrahamson,1991, 1996) seeks to unders
17、tand these dynamics. TQM re-ftects a consistent theme in recent organizational theory: aconcern that these fashions consist mainly of hype. Suchconcerns have generated considerable concern about therole of organizational theorists e.g., Beyer, 1992; Donaldson,1992). According to Astley and Zammuto (
18、1992), organiza-tional theorists and managers engage in separate “languagegames.“ Managers generate rhetoric, organizational theo-rists generate theory, and the two products cannot be recon-ciled (Astley and Zammuto, 1992). According to the researchon management fashion, ongoing relationships betwee
19、nfashion setters and fashion users fuel a demand for manage-rial fashions. Consequently, the content of the fads appar-ently means less than the value of maintaining an appear-ance of rationality and remaining at the forefront ofmanagerial thought (Abrahamson, 1996). recognize that there is some leg
20、itimatedispute over the way that Astley andZammuto (1992) have used Wittgen-steins 0958) term “language game,“ butI am interested tn the notion of “(an-guage games“ as Astley and Zammuto11992) used the term: as stylized dis-course specrfic to the semi-autonomous(gsnizatlon science and manageriaf com
21、-munities. See Mauws and Phillips (1995)for another view of Wittgensteins idea oflanguage games.Here, I am interested in TQM as one such problem, specifi-cally in how it is that a reasonably well-defined and estab-lished technical intervention like TQM can become an am-biguous and sometimes dubious
22、intervention (Hackman andWageman, 1995). It is almost as if there are two versions ofTQM. Qne TQM, a technical TQM, incorporates some fairlywell-defined organizational interventions that have clear rulesfor the use and analysis of information. A second TQM, arhetorical TQM, seems to carry the sort o
23、f rhetorical excessthat Hackman and Wageman (1995) worried about. As Hack-man and Wageman observed, from the original statisticalideas of Deming (1986) and Juran (1974), the rhetorical TQMhas exploded into a broadly used, ambiguous term with un-clear organizational implicationssave that it presumabl
24、y im-proves an organization. A distinction between technical TQMand rhetorical TQM recalls a longstanding claim of institu-tional theory (Powell and DiMaggio, 1991). Institutionaltheory describes a process whereby the symbolic value ofsomething like TQM ultimately supplants its technical (effi-cienc
25、y) value (DiMaggio and Powell, 1983). Selznick (1957:17) laid out the general thrust of the institutional argument inhis classic statement that “ to institutionalize is to infusewith value beyond the technical requirements of the task athand.“ Accordingly, the theory would suggest that the rhe-toric
26、al excess pervading TQM and similar fads follows fromthe tension between the true technical merit of the practiceand the institutional reality of its use. TQM gains institutionalvalue over time because it becomes the accepted way ofdoing things. Using TQM may provide an organization withlittle techn
27、ical benefit, but the claim to use TQM confers le-gitimacy on the organization (Westphal, Gulati, and Shortell,1997). Consequently, managers will use the rhetorical TQMto gain legitimacy without affecting activities at the technicalcore of the organization (Meyer and Rowan, 1977).603/ASQ, September
28、1998Seiznick (1996: 273) made the samepoint atxjut institutional theorv- See Scott i1995: 152) for the progressionof the distinction.Though the institutional perspective offers a provocativeview of TOM, an institutional story about TOM presents twosignificant puzzies. The first puzzie is about the r
29、hetoricaiTOM. Institutional theorists argue that the mimetic pro-cesses that fuel institutional conformity are more powerfuithan “any concrete evidence that the adopted modeis en-hance efficiency“ (DiMaggio and Poweii, 1991: 70). institu-tionalists wouid argue that in the case of rhetorical TOM, the
30、managerial rhetoric about TOM serves mostly symbolic pur-poses (Pfeffer, 1981), suggesting that managers engage inlanguage games (Astley and Zammuto, 1992). Yet the dis-pute over TOM reflects differing views about the content ofTOM. Both the quaiity gurus and the TOM advocates in or-ganizations prea
31、ch a need for a fundamentally different andbetter way of doing business. W. Edwards Deming (asquoted in Senge, 1992) claimed that TOM is a vacuous term,mere hype that perverts his work. But the executives claimthat management scholars have failed to grasp adequatelythe magnitude of TOM. To them, TOM
32、 consists of an ex-traordinary development originating in business practice(Robinson et a!., 1991). Moreover, they point to outcomes intheir own organizations that demonstrate these claims. Theylament the lack of attention given to TOM by academics.The second puzzie is about the technicai TOM. From
33、an in-stitutionai view, later adopters who choose TOM for institu-tionat reasons, rather than technicat reasons, choose to intro-duce a more narrowiy conceived TOM (Westphat, Gutati, andShorteti, 1997). According to the institutionat argument, eariyadopters wiit pursue TOM for its technicat merit an
34、d, so, wittimplement wfiatever works for them. Later adopters, how-ever, will choose TOM for legitimacy reasons and, so, willimplement a standard, and therefore presumably legitimate,TOM package. That institutionat view, however, seems atodds with TOM both theoretically and empirically. Theoreti-cai
35、iy, given the importance of operational definitions in TOMand related prescriptions (e.g., Ishikawa, 1985; Deming,1986), such increasingly narrow definitions of TOM wouidsuggest a carefuiiy deiineated operationai definition and thusseem like a good thing. Yet empiricaiiy, the evidence sug-gests just
36、 the reverse. The history of TOM shows it shiftingfrom its origins in a narrowly defined set of technical inter-ventions (Tachiki, 1992) into an extraordinarily diffuse fad,characterized by rhetorical excess and increasingly uncleardefinition (Hackman and Wageman, 1995). if managerschoose TOM to gai
37、n tegitimacy, their cotiective efforts atfinding legitimacy somehow have expanded, rather than con-tracted, the definition of TOM.These puzzles presented by institutional theory as applied toTOM reflect an ongoing ambiguity in institutionat theory. Theinstitutional view offers a powerful insight, th
38、at the adoptionof TOM might simply be “myth and ceremony“ (Meyer andRowan, 1977). Yet institutional theory says iittie about theprocess by which a managerial innovation develops from aclear technical intervention into a mere symbol of institu-tional legitimacy; the relationship between the technicat
39、 andinstitutional merits of a practice like TOM remains a sourceof considerable confusion in institutionat theory. Moreover,institutional theory continues to explain institutional conform-604/ASa September 1998Reiriltyity outside the individual organization. As DiMaggio and Pow-ell (1991: 81 argued,
40、 institutional explanations “cannot be re-duced to aggregations or direct consequences of individualsattributes or motives“ (see aiso Selznick, 1996). In supportof that ciaim, scholars have amassed considerable evidenceon the organizational adoption of institutional practices thatdemonstrate the pre
41、sence of institutional forces (e.g., Tol-bert and Zucker, 1983; Westphal, Gulati, and Shorten, 1997),Yet for all the discussion of institutional forces, institutionaltheorists have very little to say about how those forces de-velop or how those forces affect behavior in an organization.It seems puzz
42、ling that a theory rooted in the social construc-tion of reality (Berger and Luckmann, 1966) should say sotittle about how institutional value replaces technical value.Here, I consider the adoption, use, and retention of TQM ina variety of organizations to see how institutional processesshape the te
43、chnical reality of TQM. Following the logic ofthe social construction of reality, I treat TQM as a “sector“of reality (Berger and Luckmann, 1966: 24), an aspect of ev-eryday reality that has particular meanings and experiencesthat shift organizational members out of their ordinary inter-pretations o
44、f everyday reality. Individuals who encounterTQM must integrate their understanding of the technical di-mensions of TQM with the everyday realities they encounterin ongoing organizational processes. I treat the statisticalbase of TQM as a very specific sector of reality: a “provinceof meaning“ Berge
45、r and Luckmann, 1966: 25), a way ofviewing everyday reality that provides specific circumscribedmeanings to particular circumstances. To shed light on theseunderstandings that organizational members create whenthey encounter TQM, 1 studied the relationship betweenwhat people saythe rhetoric of TQM u
46、seand whatpeople dothe reality of TQM use. In this context, I havedefined rhetoric as the managers stated claims and ac-counts of TQM use, especially in the context of ongoing or-ganizational life. I have defined reality as the specific ele-ments of TQM in practice and discourse, but I ground thatre
47、ality in a set of specific technical practicesfor example,specific observable instances of TQM interventions, such asstatistical process control, data analysis tools, and idea gen-eration tools like brainstorming (Juran, 1974; Ishikawa, 1985;Deming, 1986; Hackman and Wageman, 1995). I selectedthese
48、narrow definitions because I wanted to follow the or-ganizational processes that shift TQM from the technicalTQM to a more institutional TQM as managers and organiza-tional members generate meanings out of their use of TQMand hence define for themselves the reality of TQM. tn se-lecting these narrow
49、 definitions, I took into account the longhistory and deep theoretical base of TQM (Tachiki, 1992;Hackman and Wageman, 1995), largely based in statisticalthinking and a variety of technical tools that have emergedaround that statistical base. Defining specific elements ofTQM as provinces of meaning by no means makes themmore or less real than other dimensions of a TQM program.1 could have chosen any one of a variety of different dimen-sions of TQM to define a basefor example, the smallgroups that also make up an integral part of TQM, somestrategic planning pro
