Management fashion is also shaped by its own internal dynamics.

women or children: dresses (Barthes, 1983;Richardson ; Kroeber, 1940), interior design, cooking, or children's names and toys (Lauer & Lauer, 1981).Second, theories of fashion in aesthetic forms are used unmodified to explain fashions in technical forms, such as management techniques. The author would like to thank Barbara Czarniawska-Joerges in particular for encouraging and helpful comments on earlier drafts of this article.

Thanks also to three anonymous AMR reviewers, Warren Boeker, Anders Forsell, Donald Hambrick, Bernward Joerges, Jim Kuhn, Rita McGrath, Kerstin Sahlin-Anderson, Tony Spibey, Richard Rottenburg, Guje Sevon, Pamela Tolbert, Michael Tushman, and Ruth Wageman. 54 This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 255 These theories of management fashion suggest that organizational performance gaps opened by technical and economic environmental changes do not shape the demand for management fashions; sociopsychological forces do instead.

In another article, for instance, I adopted Blumer's (1969) argument that fashion demand is guided by managers" collective aesthetic tastes (Abrahamson, 1991).Likewise, Mintzberg (1979) observed that swings between organizational centralization and decentralization resemble the movements of women's hemlines. The popular management press is even more cynical and strident, attributing the demand for management fashions to other sociopsychological forces, such as childlike excitement (Business Week, 1986), mass conformity (Wall Street Journal, 1993),and even to something akin to manias or episodes of mass hysteria (HarvardBusiness Review, 1994).These academic and popular sociopsychological treatments of management fashion cast it as something from which management scholars should remain disengaged, lest they join the ranks of "snake-oil salesmen. " I argue in this article that management fashion should not be treated as

a special case of aesthetic fashion. Aesthetic and management fashions differ in two important ways, and failing to recognize these differences obscures why management scholars must not only study, but also intervene in the management-fashion-setting process.

First, whereas aesthetic fashions need only appear beautiful and modern, fashionable management techniques must appear both rational (efficient means to important ends) and progressive (new as well as improved relative to older management techniques). Many management fashion setters-consulting firms, management gurus, business mass-media publications, and business schools-compete in a race to define which management techniques lead rational management progress.Fashion setters who do not participate successfully in this race, business schools and professional scholarly societies, for example (Hambrick, 1994), will be perceived as lagging rather than leading management progress, as being peripheral to the business community, and as being undeserving of societal support. Hence, this article warns that scholars in business schools must both study and intervene in the management-fashion-setting process; otherwise these business schools' long-term viability will be at risk.Swings in management fashion, far from being cosmetic and trivial, are in fact deadly serious matters for business schools and the scholars staffing them. Second, I argue that whereas sociopsychological forces alone shape the demand for aesthetic fashions, such forces compete with technical and economic forces to shape the demand for management fashions.

Put differently, managers do not adopt management fashions only because of sociopsychological forces.They also adopt management fashions in a desire to learn about management techniques that would help them respond to organizational performance gaps opened up by real technical and economic environmental changes. Management fashion setting, consequently, can serve as a technical learning process for many managers. This content

downloaded from 149.171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 256 Academy of Management Review January Business schools can, and indeed must, study and intervene in this collective learning process, in order to render it more technically functional.The first of this article's three parts defines the term management fashion in a way that emphasizes how it differs from aesthetic fashion.

The second part advances a theory of management fashion, drawing principally on two literatures in organizational theory: the neoinstitutional theory literature (Meyer & Rowan, 1977; Scott & Meyer, 1994) and the production of culture literature (Blau, 1993; DiMaggio & Hirsch, 1976; Hirsch, 1972, 1986; Peterson, 1976, 1979). The third part uses this theory of management fashion as a framework to review a variety of prescriptions advanced by anagement scholars, for management scholars interested in influencing management practice (Astley & Zammuto, 1992; Bettis, 1991; Beyer, 1992; Donaldson, 1992; Nystrom & Starbuck, 1977; Starbuck & Nystrom, 1981). The article closes with a plea to management scholars not only to study the management-fashion-setting process and explain when and how it fails to serve the needs of shareholders, employees, managers, students, and other business school stakeholders, but also to intervene in this fashion-setting process, in order to make it a more technically useful collective learning process for these stakeholders.

Management Redefining Fashion

We experience fashions as rapid, bell-shaped swings in the popularity of management techniques.

This alone, however, cannot serve as a definition, because a number of processes could cause such swings. ' For this reason, I argued in an earlier article that we should label these swings "management fashions" only when they

are the product of a management-fashion-setting process involving particular management fashion setters-organizations and individuals who dedicate themselves to producing and disseminating management knowledge (Abrahamson, 1991).My earlier treatment of management fashion as both a process (fashion setting) and its outcomes (management fashions) is useful here in defining management fashion. It does not explain, however, why many managers would gain and lose interest in management techniques launched by management fashion setters. In this article, I argue that waves of interest in management techniques occur when national norms of both rationality and progress govern managerial behavior.

Norms of managerial rationality are societal expectations that managers will use A swing in the popularity of a management technique might occur, for example, if a government regulator mandated the use of a management technique and then withdrew this mandate. Alternatively, a popularity swing might occur because a management technique that gained popularity because it helped narrow a performance gap caused by an environmental change lost this popularity when the environment changed again, narrowing this performance gap. This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 257 anagement techniques that are the most efficient means to important ends.

Norms of managerial progress are societal expectations that, over time, managers will use new and improved management techniques. Together, norms of managerial rationality and progress create the need for a flow of management techniques that organizational stakeholders believe are rational, at the forefront of management progress, and that managers can adopt in order to appear in conformity with these norms. The management-fashion-setting process produces the continuous flow of management techniques

believed to be rational and progressive.It is for this reason that I define management fashion setting as the process by which management fashion setters continuously redefine both theirs and fashion followers' collective beliefs about which management techniques lead rational management progress. These shared beliefs about management progress cannot remain stable for too long; otherwise, progress will not seem to occur. A management fashion, therefore, is a relatively transitory collective belief, disseminated by management fashion setters, that a management technique leads rational management progress.

This definition suggests that management fashions can differ in scope: The collectivity that believes that a management technique is at the forefrontof management progress can be large or small. Management fashions also can vary in duration: The belief that a management technique is at the forefront of management progress can be more or less transitory. Quality circles (QCs) exemplify management fashion. During the early 1980s, fashion setters promoted the transient belief that QCs were at the forefront of management progress.The rhetoric that fashion setters used to promote this belief survives in popular management press articles, as well as in the proceedings of meetings of fashion setters who actively promoted the QC fashion.

An examination of the number of these articles and the size of these proceedings volumes reveals the transient backing that fashion setters gave to QCs. Figure 1 indicates that the number of articles on QCs listed yearly in the ABIInformdatabase grew rapidly in the early 1980s,yet this trend was reversed after 1983.By 1986, this measure returned to its prepopularity level, reflecting the swing in QCs' popularity. The International Association of Quality Circles (IAQC),an association of practitioners and consultants, meets yearly and publishes

the proceedings of these meetings.

Figure 1 reveals a bell-shaped pattern in the thickness of the yearly IAQC proceedings volumes. Surveys reveal that the bell-shaped popularity pattern of QCs among fashion setters also occurred among fashion followers. Nearly one third of U. S. rganizations with more than 500 employees reported adopting QCs between 1980 and 1982 (New York Stock Exchange, 1982).

Lawler and Mohrman (1985) estimated that 90% of the Fortune 500 companies had adopted QCs during this period. At the same time, they also predicted an This content downloaded from 149. 171. 67.

164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 258 Academy of Management Review January FIGURE 1 Print-Media Indicators of Quality Circles 45-. 4035 30- 1015- 0- I I III 8 77 78 79 80 81 82 YEAR 83 84 85 86 87 _1EW_ Number of articles -Width of International Association of Quality Circles Proceedings in decimeters. impending decline in the use of QCs. A survey by Castorina and Wood (1988) indicated that more than 80% of Fortune 500 companies that adopted QCs in the 1980s had abandoned them by 1987.

Evidence in the print media also indicated that management fashions may not be a recent phenomenon. Certain management techniques have repeatedly gained and lost popularity since the turn of the century. Figure 2, for example, reveals three swings in the number of articles about employee stock ownership published between 1914 and 1986.Certain obstacles still block careful theorizing about management fashions. Researchers often implicitly believe that management fashions hold sway only in relatively trivial management matters. This implicit belief exists because words like fashion and fad, because of

their previous uses in primarily aesthetic realms, now connote the terms unimportant and trivial, when they are used in technical realms, such as management.

It is this article's contention that such an a priori assumption may be unfounded (Czarniawska-Joerges & Joerges, 1990).More generally, management researchers should remain open to the possibility that This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 259 FIGURE 2 Prevalence in the Number of Articles on Employee-Stock-Ownership Programs 30 2826 24 22 20 18Z 16141210 86- o 42- 0 14 22 30 38 46 54 62 70 78 86 YEAR not only management, but also marketing, finance, accounting, operations, as well as almost every area of technical endeavor are open to the swings of fashion.Theorists should therefore attempt to develop theories of fashion in technical realms, without assuming outright that these theories will explain fashions only in either aesthetic matters or in trivial technical matters.

In the next section, I develop such a theory of management fashion in three parts, which correspond to the three parts of Figure 3. First, I develop the argument that management fashions are about the appearance of rationality and progress. Here, I draw primarily on the neoinstitutional theory literature (Meyer & Rowan, 1977; Scott & Meyer, 1994).I suggest, in accordance with the top part of Figure 3, that in certain countries, norms of rationality and progress create a management fashion market for rhetorics2 championing rational and progressive management techniques. In the second part of this section, I draw on the production of culture literature to conceptualize the mutual influence, depicted 2

A rhetoric is spoken and written discourse that justifies the use of a set of techniques for managing organizations or their employees (Barley & Kunda, 1992). This content downloaded from 149.

171. 67. 64 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 260 Academy of Management Review January FIGURE 3 A General Model of Management Fashion Setting Norms of Rationality and Progress Management Fashion Market Supply by Management Fashion Setters Demandby Management Fashion Users |Sociopsychological and Technoeconomic Forces in the middle of Figure 3, between the management-fashion-setting community that supplies these rhetorics and the management-fashion-using community that demands them (Blau, 1993; DiMaggio & Hirsch, 1976; Hirsch, 1972, 1986; Peterson, 1976, 1979).This focus on supply and demand can reveal how business school scholars might better intervene in management fashion setting. In the third part, I examine forces depicted at the bottom of Figure 3, which are external to the management-fashion-setting market, but nevertheless influence management fashion demand.

A focus on such forces helps to explain the timing of management fashion swings, the direction of management fashion trends, and when business school scholars might effectively intervene in the shaping of management fashion. This content downloaded from 149. 171. 67. 64 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 261

A Theory of Managementfashion

Norms of Rationality and Progress My preliminary sketch of management fashion suggests that it is largely a cultural phenomenon, shaped by norms of rationality and progress.

Therefore, my theory of fashion draws and builds upon one perspective that focuses on such normative influences: neoinstitutional theory. More specifically, there currently exists many variants of

neoinstitutional theory. (See Scott, 1987; Powell & DiMaggio, 1991, for reviews. Therefore, I concentrate on Meyer and Rowan's (1977) variant, because of its focus on norms of rationality (See also Meyer & Scott, 1992; Scott & Meyer, 1994.

) Norms of rationality. Meyer and Rowan (1977) asserted that organizational stakeholders expect managers to manage their organizations and employees rationally-that is, by the most efficient means to important ends. In many contexts, however, it is ambiguous not only what constitutes important ends for managers to pursue, but also what are the most efficient means to pursue these ends (March & Olsen, 1976).Under these circumstances, managers must create the appearance that they are conforming to norms of rationality.

Meyer and Rowan (1977) suggested that managers create the appearance of rationality by using or appearing to use management techniques that generally are believed by organizational stakeholders in a specific context to be rational ways of managing organizations and employees. They describe such rational management techniques as "prescriptions that identify various social purposes as technical ones and specify ...

he appropriate means to pursue these technical purposes rationally" (Meyer & Rowan, 1977: 343-344). In other words, rational management techniques are labels that denote for organizational stakeholders both certain managerial goals that effective managers should pursue, as well as the means to pursue these goals efficiently. The label quality circle, for example, denotes both the pursuit of product quality as a goal of paramount importance, and the means to attain this goal efficiently: organizing meetings of workers during which they discuss among themselves ways of enhancing product quality.Managers appear rational if they appear to use management techniques, such as quality circles, that stakeholders

believe are efficient means to important ends. Meyer and Rowan (1977) argued that if managers do not appear to use such techniques, then stakeholders' expectations that the organization is run rationally will tend to be disappointed, and stakeholders will tend to withdraw their support from the organization, the thereby increasing likelihood that this organization and its managers will fail.

It is not a straightforward task to extend Meyer and Rowan's (1977) argument to incorporate the notion that not only norms of rationality, but also norms of progress govern managerial behavior. It is straightforward that if, (a) under conditions of ambiguity, norms of rationality cause managers to use management techniques perceived as rational in order to This content downloaded from 149. 171. 67.

164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 262 Academy of Management Review January ustain the appearance of rational management and to avoid stakeholder sanctions, then (b) under conditions of ambiguity, norms of progress should cause managers to adopt management techniques perceived as progressive in order to create the appearance of management progress and to avoid such sanctions. Two aspects of this argument are not straightforward. First, it is unclear what defined so-called norms of progress. Second, it is also unclear how management techniques come to be considered both rational and progressive, rather than irrational and retrogressive, that is, as violations of tried and true rules of rational management.

Norms of progress. In this article, I define a management innovation as "a significant departure from the state of the art at the time it first appears" (Kimberly, 1981: 86). Not all innovations are improvements. Therefore, I use the

term progress to mean both innovation and improvement as judged by an evolving set of criteria.

This definition evolves from the notion that norms of progress do not generate expectations of progress judged against some invariant end or utopia (Lasch, 1991).Rather, they generate expectations of a never-ending improvement process judged by criteria that are repeatedly redefined in and by this process itself. Norms of scientific progress are the exemplar: They create expectations of greater understanding according to theoretical and methodological criteria repeatedly redefined by the scientific process itself. In direct parallel, norms of management progress create an expectation of ever-improved techniques for managing organizations and individuals, judged by everimproving criteria for judging what constitutes managerial improvements.These criteria, be they quality, flexibility, profits, speed, client satisfaction, or risk reduction, are inherently subjective and are neither absolute, fixed, nor universal (Granovetter, 1979). It also cannot be assumed that norms of management progress are themselves universal.

Indeed, research on national cultural values sugequal strength across national cultures, suggesting an initial test of this article's theory of management fashion (Hofstede, 1980; Inkeles & Smith, 1974). 3 Management fashion markets probably differ across countries in More specifically, Kluckhohn and Strodtbeck's (1961)classical typology of national cultures contains two dimensions-human-nature orientation and value orientation-that may be particularly telling in determining in which nations strong norms of management progress will be found and management fashions will tend to hold sway. The human-nature dimension distinguishes national cultures with respect to their assumptions about whether or not individuals are adaptable or perfectible. The time-orientation dimension distinguishes national cultures that focus on the future, the present, or the past.

The culture of the U. S. , for example, assumes

that individuals are adaptable to a dynamic, changing future. This cultural orientation may explain U.

S. culture's strong norms of progress, as well as the large market of the United States for self-help books advocating progressive techniques for managing not only mental and physical health, but also organizations and their employees. By contrast, norms of progress may tend to be weaker in national cultures that assume that individuals are either not perfectible or perfectible primarilyalong paths dictated by customThis content downloaded from 149. 171.67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 263 both the frequency and duration of management fashions. If norms of management rationality and progress drive the market for management fashions, then Proposition 1 follows: Proposition 1: There will tend to be more frequent and shorter lived management fashions in nations that have relatively stronger norms of rational management progress. Deviance or progress?The use of apparently new management techniques can make their users appear irrational and retrogressive (violators of tried and true rules of rational management) to organizational stakeholders, rather than rational and progressive (Perrow, 1970; Starbuck & Nystrom, 1981).

This raises the following question: If norms of progress call for a flow of apparently rational and progressive techniques to sustain the appearance of continuing rational progress, then how do organizational stakeholders come to perceive these techniques as rational and progressive rather than as irrational and retrogressive? ommunity My answer is that there is a management-fashion-setting fashion setters. It is this managementmanagement populated by community that shapes transitory collective beliefs fashion-setting among management fashion followers that certain management techniques are rational and at the

forefront of management progress. Put differently, management fashions do not emerge spontaneously as a result of the inventive behaviors of managers.They are cultural commodities deliberately produced by fashion setters in order to be marketed to fashion followers. I turn next to a body of theory and research uniquely suited to studying the production of management fashions: the literature on the production of culture (Peterson, 1976, 1979). The Production of Management Culture As Peterson wrote, "The production of culture perspective begins with the observation that symbolic elements of culture do not spring forth full blown but are made somewhere by someone" (1979: 152).

Theories and studies of fashion from this perspective have gone a long way toward dispelling the notion that cultural innovations become mass fashions by direct popular demand (Blau, 1993; Blumer, 1968, 1969; DiMaggio, 1982, 1987; DiMaggio & Hirsch, 1976; Hirsch, 1972, 1986; Peterson & Berger, 1975). Entire industries often stand between the creators of innovations and the masses who use these innovations if they become fashionable. These industries produce the cultural fashions that the masses consume. In this section, I generalize the theory of sartorial, literary, musical, and and traditions.Weaker norms of management weaker market for management fashions progress would create a commensurably This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 264 Academy of Management Review January cinematographic fashion found in the production of culture literature to the realm of management fashion.

Hirsch (1972), in particular, focused on organizations in the publishing, record, and motion picture industries that mediate between, on the one hand, writers, musicians, or actors, and on

the other hand, consumers of literary, musical, or cinematographic mass fashions.He described a four-fold fashion-setting process (see also, DiMaggio & Hirsch, 1976). First, cultural innovations are created in show-business, musical, or literary circles. Second, fashion-setting organizations use talent scouts to penetrate these artistic circles and to discover new talents and cultural innovations. Talent scouts select a small set of manuscripts, scores, or scripts from the vast array of literary, musical, and cinematographic innovations.

Third, fashion setters process these innovations with an eye to the market to which they are targeted. Manuscripts are edited and published.Scores are refined and recorded. Scripts are rewritten and enacted.

Fourth, fashion setters attempt to disseminate the cultural products that they have processed. They advertise these products and sometimes go so far as to attempt to co-opt mass-media gatekeepers (critics, editors, etc. ), in order to bring these products to the public's attention. As a result of this four-fold process (creation, selection, processing, and dissemination), fashion-setting industries supply mass audiences with a limited set of innovations that are candidates for becoming mass fashions.

These innovations may or may not become mass fashions, depending on fashion followers' demand for these innovations (Blumer, 1968, 1969; DiMaggio & Hirsch, 1976). Members of organizations that set fashions, therefore, either thrive or falter, depending on their ability to sense and satiate incipient preferences shaping fashion demand for cultural innovations. In line with Hirsch's (1972) model, Figure 4 depicts the creation, selection, processing, and dissemination by fashion suppliers of rhetorics championing certain management techniques.In the United States, various scholars have claimed that a variety of organizations and individuals populate a management-fashion-setting community: management consultants, business schools, and business-press organizations (Abrahamson,

1991; DiMaggio & Powell, 1983; Hirsch, 1986; Kimberly, 1981; Meyer, 1992; Mintzberg, 1979; Strang & Meyer, 1994) as well as academic gurus, consultant gurus, and hero managers (Huczynski, 1993).

The bottom box in Figure 4 depicts the creation, selection, processing, and dissemination of management fashions by management fashion suppliers.The arrow leading out of the right bubble indicates that during the creation stage, fashion setters sense incipient preferences guiding fashion demand and create many management techniques. During the selection stage, they select those techniques that they believe will satiate this demand. The arrow leading out of the left bubble suggests that during the processing stage, fashion setters articulate rhetorics championing the management techniques they select.

During the dissemination stage, This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 265FIGURE 4 ashion-Setting Management Fashion Market ..

This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 266 Academy of Management Review January involve either inventing management techniques that only appear to be old management techniques improvements or rediscovering/reinventing that were invented previously and forgotten.Many critics have noted this "old wine in new bottles" phenomenon (e. g. , Kimberly, 1981). Most management innovations may be created by managers. Galbraith (1980: 162) asserted: I know of no new form of organization that was invented by organization theorists while advancing the theory. I have seen no new formemerge fromthe test tubes of organization theory. Instead, the researchers record what the inventive practitioner creates and give it labels like grids, system 4, or matrix organization.

Alternatively, certain fashion setters may invent, rediscover, or reinvent the management technique they attempt to launch into fashion.W. Edwards Demming and others are frequently credited with the invention of the TQM approach, and BCG consultants have been credited with the BCG matrix, for example. Selection. Virtually nothing is known about the selection stage of management fashion supply. It is reasonable to assume that fashion setters can always select the techniques they attempt to launch into fashion from a broad variety of new, recently created management techniques, as well as from a bevy of old, forgotten management techniques.Either or both of two very distinct selection processes might operate, however, depending on whether it is managers or fashion setters who create management innovations. One selection process resembles the process described by Hirsch (1972), in which artists invent cultural innovations in artistic circles located outside the fashion-setting community, and fashion setters scout out such innovations, select a few, and bring them into the fashionsetting community for processing and dissemination.Likewise, management innovations may be invented by managers and selected into the community by management fashion setters management-fashion-setting into contact with these managers. who come A second selection process also could operate, if it is fashion setters who either create management innovations or resurrect old and forgotten management techniques.

Fashion setters may select among techniques they themselves or other fashion setters have invented or rediscovered. Other management fashion setters may then imitate these selection decisions.Why do management fashion setters select certain management techniques and not others? The answer is twofold and dynamic: Fashion setters not only (a) sense and satiate incipient demand for new types of management fashions, but they also

(b) shape and focus this demand by articulating for fashion followers the particular techniques that fit the types followers prefer. I now examine each process in turn. This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 996 Abrahamson 267 First, as the bottom of Figure 4 indicates, and as I will argue in the next subsection, new sociopsychological and technoeconomic forces excreate new demand for particular ternal to the fashion-setting-market types of management techniques. Global competition heats up; for example, customers become more sensitive to product or service quality, and managers seek out new techniques that could help them deal with performance gaps caused by these environmental changes.Initially, fashion setters sense only dimly the incipient demand for new fashions and struggle to select management techniques that they believe might satiate this incipient demand. Second, when management fashion setters select and disseminate these techniques, some become fashionable and diffuse. These fashions shape fashion users' incipient preferences, guiding demand for upcoming fashions. These fashions also clarify somewhat for fashion setters both the type of management fashions in demand and what the next fashion they select should look like.A nonmanagerial innovation illustrates this process. When some customers responded positively to cars with rounded bodies in the 1980s and purchased them, they reinforced other customers' tastes for a more rounded car silhouette. This positive response also alerted car designers that the following year's model would sell if they selected an even more rounded silhouette, triggering the ongoing fashion trend toward ever more rounded car bodies. 4 Like car bodies, management techniques that become fashionable may

further focus managers' preferences, generating preferences for related management fashions.This causes fashion setters to recognize and select related management fashions, causing a management fashion trend. The quality circle fashion, for instance, can be seen as a precursor of a number of related management fashions belonging to a total quality fashion trend (Dean ; Bowen, 1994).

In sum, it is important to remember that there exists a reciprocal relation between what fashion setters select and what fashion consumers prefer and demand (Blumer, 1968, 1969). Processing.When a management technique is selected by fashion setters, in what sense do they process it? Processing involves the elaboration of a rhetoric that can convince fashion followers that a management technique is both rational and at the forefront of management progress. How does a rhetoric create the belief that a management technique is rational? It must create the belief that the technique allows managers to pursue an important managerial goal in the most efficient fashion.It must, therefore, articulate (a) why it is imperative for managers to pursue 4 Of course, there is a limit to just how rounded a car body can become, just as there are limits to the length or shortness of hemlines. Robinson (1958) argued that when fashion trends cannot progress any further, fashions switch swiftly and abruptly to the midpoint of another dimension, along which a new fashion trend can proceed (rounded car bodies become increasingly wider, for example). This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 68 Academy of Management Review January this goal and (b) why this technique provides the most

efficient means to attain these goals. I examine, in turn, how rhetorics achieve each of these dual objectives. Rhetorics create the belief that managers must pursue certain goals by highlighting organizational performance gaps whose goal it should be for managers to narrow. Rhetorics do this in different ways. Some rhetorics present a gap with great drama and fanfare, in order to highlight why narrowing such a gap is an effective managerial goal.

They attempt to frighten managers by revealing threats of managerial demise if such a performance gap is ignored, and to entice them with sublime opportunities for managerial stardom if this gap is attended to (Chen ; Meindl, 1991). Other rhetorics discuss careful scientific evidence bearing on the existence and importance of such performance gaps. The rhetoric also must create the belief that the technique it champions is the most efficient means to attaining an important goal.Arguments run the gamut from (a) descriptions of techniques whose use by a few currently successful companies is used to justify claims about the efficiency of these techniques in narrowing performance gaps for all companies, to (b) quasi-theoretical statements of how these techniques cause performance gaps to narrow, to (c) full-fledged empirically validated, scientific theories of the causes of such effects. Arguments of the first type have the following form: Quality circles were used at the time of the Japanese miracle; therefore, implant them in any and all U. S. firms, and watch the U.S. miracle happen. Scientific grounding, usually provided by citations or miscitations of business school research, is sometimes added to give these rhetorics a patina of rationality. At the other extreme, arguments of the third type

are carefully grounded in scientific research produced by business schools. The job-enrichment fashion, for example, was backed by a solid body of empirical research produced by business school scholars. Rhetorics must not only create the belief that the techniques they champion are rational, but also that they are at the forefront of management progress.Rhetorics vary in how they achieve this end. Certain rhetorics clearly indicate how a management technique constitutes an innovation, or reveal the history of successes and failures of a management innovation being brought back into fashion. These rhetorics also disclose scientific evidence, indicating that the management technique provides an improvement over techniques currently in use, as judged by carefully articulated criteria. Other management rhetorics provide no careful evidence that techniques constitute improvements.Moreover, they present old and forgotten management goals and innovations as if they had just been invented. Dissemination. We know little about the dissemination of progressive management rhetorics from management fashion setters to users. Hirsch (1972: 643) argued that we should view "the mass media in their gatekeeping role as a primary institutional regulator of innovation" and that This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 996 Abrahamson 269 "the diffusion of particular fads and fashions is either blocked or facilitated at this strategic checkpoint" (Hirsch, 1972: 649). Indeed, mass-media publications reach broad audiences and, therefore, have the potential to draw widespread attention to progressive management rhetorics championing particular management techniques. Burns and Wholey (1993), for example, found that press coverage bearing on the use of matrix forms in hospitals correlated with their adoption across hospitals.There

are two reasons, however, why it may be necessary to move to a finer grained conceptualization of mass-media effects on the dissemination of management fashions. First, a variety of publications may contribute directly to the dissemination of management fashions (Barley, Meyer, ; Gash, 1988; Dunbar, 1983).

Diverse groups of fashion setters, for example, have developed their own publication to disseminate their rhetorics: McKenzie Quarterly (consultants), Business Week (journalists), Academy of Management Executive (academics).The Academy of Management Executive, in particular, has a "Research Translations" section in each issue, in which the academic rhetoric of academic studies is translated into a rhetoric understandable to practitioners. Management gurus, moreover, use the book-publishing industry to disseminate their rhetorics (Huczynski, 1993), as do management educators use the textbook industry to disseminate their rhetorics. Therefore, it may be necessary to distinguish the impacts of different types of publications on which techniques become fashionable and diffuse.There is a second reason why a finer grained conceptualization is necessary. Whereas certain fashion setters may disseminate their rhetorics directly to managers through the publications they control, other fashion setters may do so indirectly, when publications they do not control appropriate their rhetorics. It is generally assumed, for example, that academic fashion setters exert most of their influence in such an indirect fashion (Beyer, 1992).Barley, Meyer, and Gash (1988), however, in a rare study of its kind, studied how corporate culture rhetoric spreads between the academic and popular-management press. Their study suggested that the corporate culture rhetoric, developed in the popular-management press, shaped the academic press's rhetoric, rather than vice versa. Their study raised the disquieting possibility that the popular-management press tends to lead the dissemination of

progressive-management rhetorics, whereas the academic-management press tends to lag.Alternatively, such a lag may reflect attempts by management scholars to directly influence management practitioners by using their rhetoric. Clearly, more research, like that of Barley, Meyer, and Gash (1988), which uses print-media traces to study management-fashion dissemination, is needed to establish the existence of such lags.

Moreover, more research will be required to ascertain what are the long-term consequence of these lags, if they should be reversed, and whether such a reversal is possible.My theory of fashion suggests, however, that fashion setters who appear to lag rather than lead management progress also will tend to lose their stakeholders' support. Beyer (1992: 471), for example, This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 270 Academy of Management Review January warned that an "occasional trickle of concepts [from management scholars to practitioners] hardly seems a sufficient rationale to sustain the social legitimacy and financial support of all of the faculty, Ph.D. programs, and research projects in our field. " Parallel to this line of reasoning, my theory of management fashion suggests that fashion setters who tend to lag in the dissemination of progressive management rhetorics also will tend to be perceived as lagging behind management progress and, consequently, as undeserving of their stakeholders' support. In this section, I began by discussing how norms of rationality and progress open up markets for management fashions.I then turned to a description of the fashion-setting process in this market by which management techniques are created and selected, rhetorics championing these techniques are processed, and these rhetorics are disseminated

to fashion followers in order to launch fashions. My argument suggests the following proposition. Proposition 2: If there exists unmet demand for a certain type of management technique, then one or more techniques belonging to this type will become fashionable if they are created, selected, processed, and disseminated by the management-fashion-setting community.The literature bearing on the diffusion of innovative managerial techniques has examined both the rate and extent of their diffusions, as well as the sequencing of their adoptions (Abrahamson & Rosenkopf, 1993; Davis, 1991; Fligstein, 1985, 1991; Haveman, 1993; Knoke, 1982; Tolbert & Zucker, 1983). Strang and Meyer (1994) criticized this literature for its narrow realist focus on how diffusion is channeled by interorganizational relations, whether they be communication links, competitive rivalries, geographic proximity, or reputational orderings.They call for a more idealist focus stressing how perceptions of innovations influence their diffusions. I suggest that fashion setters create the perceptions that new and improved rational techniques exist, thus rendering them fashionable, and promoting their rapid and widespread diffusions across interorganizational networks. Thus, I propose Proposition 3: When a management technique becomes fashionable in a collectivity, it will tend to diffuse rapidly and extensively across organizations in this collectivity.Next, I examine forces, external to the management-fashion-setting market, that shape management-fashion demand, and which, therefore, influence whether techniques become fashionable and whether they diffuse.

Forces Shaping Management Fashion Demand I argued in the introduction to this article that theories stressing the effect of sociopsychological forces shaping management fashion demand This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 996 Abrahamson 271 must be supplemented

by a theory highlighting technical and economic forces shaping such demand. More generally, I contended that sociopsychological and technoeconomic forces compete to shape management fashion demand. Which of these forces prevail depends, in part, on whether and how management scholars intervene in the fashion-setting process. For these reasons, I begin by extending sociopsychological theories of aesthetic fashion demand to the realm of management fashion.These theories assume that each fashion follower in a collectivity is rendered vulnerable to fashion setters by one of three psychological states that I review in turn: (a) frustration, (b) boredom and striving for novelty, and (c) striving for status differentiation. In these sociopsychological theories, it is the set of managers who experience a particular psychological state that form the collectivity among which a fashion can propagate. I then advance a technoeconomic theory of management-fashion demand.It suggests that a variety of (a) economic, (b) political, and (c) organizational forces open gaps between organizations' actual and desired performances. The management-fashion-setting process brings these performance gaps to collective awareness and articulates new progressive and collectively acceptable techniques for narrowing these gaps. In this technoeconomic fashion, it is the set of managers who experience a particular performance gap that form the collectivity among which a fashion can propagate. Sociopsychological explanations of fashion demand.I have argued that norms of rationality and progress create a continuing demand for management fashions to sustain the impression of management progress. A broader variety of other forces also must be considered, however, because normative influences alone quite likely cannot sustain management fashion demand.

A variety of explanations suggests that fashions also are in demand because they satiate individuals' psychological needs. These explanations

are both the oldest (Bergler, 1953; Fluegel, 1930; Hurlock, 1929) and the ones most recently found reflected in current treatments of management fashion aimed at nonacademic audiences (e. . , Business Week, 1986). A classic article by Sapir (1937) provided a dated, yet outstanding and succinct, example of this type of explanation that may be particularly useful in explaining management fashion demand. He suggested that fashions gratify competing psychological drives for individuality and novelty, on one hand, and conformity and traditionalism, on the other. Applying this explanation to the realm of management fashion suggests that managers demand management fashions to appear individualistic and novel, relative to the mass of managers who are out of fashion.They maintain some measure of conformity and traditionalism, however, by using techniques used by other managers who are in fashion. However, what is new and individualistic ineluctably becomes old and common with the passage of time and with the growing number of adherents to a This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 272 Academy of Management Review January fashion. This explanation suggests, therefore, that management fashions will always be in demand to feed managers' insatiable appetites for individuality and novelty.Thus, Proposition 4: Management fashions satiate fashion followers' psychological desire to appear individualistic and progressive, without appearing deviant and retrogressive. A second type of sociopsychological explanation suggests that frustration and despair across individuals in a collectivity loosens normal institutional constraints on their behavior and renders them vulnerable to unrealistic hopes that quasi-magical solutions will relieve their sources of frustration (Gill & Whittle, 1992; Klapp, 1969; Smelser,

1962).An unrealistic hope, in time, can only give way to a new round of frustration and despair, and a receptivity for a new hope-inducing fashion. Versions of this argument abound in treatments of management fashion aimed at nonacademic audiences (Business Week, 1986; Eccles & Nohria, 1992; Harvard Business Review, 1994; Wall Street Journal, 1993). In these accounts, the Japanese competitive challenge, as well as the relative decline of U. S. and European businesses, generated a powerful source of collective frustration for U.S. and European managers, which, in turn, fueled management fashion demand (Mitroff & Mohrman, 1987).

Thus, if management fashion setters are responsive to fashion demand, Proposition 5: There will be more management fashions during periods in which managers' expectations are disappointed. For example, there should have been an increase in the number of management fashions in the U. S. at the beginning of the period of Japan's ascendancy.A classical article by Simmel (1957) suggested a third, more sociological explanation bearing on fashion demand. It suggests that fashions serve not only to reveal who is in fashion, but also to distinguish highstatus from low-status individuals. This explanation, when extended to management fashion, suggested what Abrahamson and Fombrun (1994) called a trickle-down fashion process. The managers of higher reputation organizations adopt management fashions to distinguish their organizations from lower reputation organizations.The more the managers of lower reputation organizations adopt fashionable techniques to make their organizations look like higher reputation organizations, however, the more both higher and lower reputation organizations look alike; hence, the greater the pressure on the managers of higher reputation organizations to adopt a new fashion that will redistinguish their organizations from lower reputation organizations.

Therefore, if management fashion setters are responsive to fashion demand, then Proposition 6: New management fashions will tend to emerge when old management fashions have been adopted by lower reputation organizations.This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 273 Technoeconomic explanations of fashion demand. Technical and economic forces shape management-fashion demand. This is not to say, however, that there exists a powerful technical determinism, whereby objective performance gaps opened up by real technoeconomic changes magically produce optimally technically efficient management fashions.Rather, technical and economic environmental changes create incipient preferences among fashion followers for certain types of management techniques that they find useful in narrowing performance gaps opened up by these environmental changes. However, as was noted previously in the section on the selection of management fashions, fashion setters not only sense and satiate incipient demand for new types of management fashions, but they also shape this demand by articulating for fashion followers that particular technique which matches the type they prefer.

A variety of technical-economic forces could influence fashion demand.Macroeconomic fluctuations could have this effect. Long-term, 50year Kondratieff waves of economic expansion and contraction (Kondratieff, 1926; Schumpeter, 1935) may parallel broad changes in managers' preferences for different types of management rhetorics (Abrahamson, In press; Barley ; Kunda, 1992). Barley and Kunda's (1992) review of historians' works supports the hypotheses that because profits hinge on capital investment and automation during expansionary periods, there should be a demand for types of management techniques that stress the efficient use of structures and technologies as a means of increasing labor roductivity. During periods

of contraction, however, both the supply and returns on capital investment decline, managers gain interest in labor as a factor of production, and there should be a demand for types of management techniques that stress employee relations as a means of increasing labor productivity. It follows that if management fashion setters are responsive to management fashion demand, then Proposition 7: The ebb and flow of management fashions will be related to macroeconomic fluctuations.A second class of technoeconomic explanations stresses the existence of political forces that cause changes in the demand for management fashions. Neo-Marxist accounts stress the inherent, irreconcilable, structural conflict between managers and workers. In this approach, management fashions are seen as so many forms of control that emerge to suppress this conflict. This conflict is unsolvable, however, as long as these techniques do not change the underlying structure that generates the conflict.Therefore, all management fashions eventually fail, and new fashions must emerge to resuppress the conflict.

As Edwards (1979: 18) put it, transformations in management fashions occurred as a resolution of intensifying conflict and contradiction in the firm's operations. The period of increasing tension was followed by a relatively rapid process of discovery, experimentation, and implementation, in which new systems of This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 274 Academy of Management ReviewJanuary control were substituted for the older, more primitive ones. Once instituted, these new relations tend to persist until they no longer effectively counter worker resistance. Other authors argue more simply that gaps between the actual amount of labor-union activity and the amount desired by managers

may cause them to gain interest in management techniques that allow them to narrow these gaps by forestalling the discontentment that causes employees to join labor unions and to strike (e. g. , Abrahamson, In press; Guillen, 1993; Jacoby, 1985).It follows that, if management fashion setters are responsive to management fashion demand, then Proposition 8: The ebb and flow of management fashions will be related to labor strife and labor-union activity. Another technoeconomic explanation suggests that unsolvable contradictions originating from within organizations, rather than without, cause changes in fashion demand (Blau, 1971). Centralization, for instance, allows for centralized control but stifles the autonomy of parts. Decentralization allows for such autonomy but causes a loss of centralized control.Fashions cause organizations to centralize, lose autonomy, and become receptive to new fashions causing them to gain autonomy, lose control, and become receptive to a new fashion that swings the pendulum back again. Likewise, fashions for flexibility cause organizations to lose efficiency causing receptivity for fashions that increase efficiency but reduce flexibility. Similar technical contradictions may cause pendulum swings in fashion demand for techniques stressing employee motivation versus control, product cost versus quality, and in-sourcing versus out-sourcing, to name a few.Thus, if management fashion setters are responsive to management fashion demand, then Proposition 9: The ebb and flow of management fashions will be related to technical contradictions within organizations.

A Pleato Studyand Influence Management Fashion

As I stated in the introduction, the dominant approach to conceptualizing management fashion treats management techniques as aesthetic forms and management fashion demand as resulting from social and psychological forces (Abrahamson, 1991; Blumer, 1968, 1969; Mintzberg, 1979).In this dominant approach, fashion setting is cast as a

form of "superstitious learning"; that is, "the subjective experience of learning is compelling, but the connection[s] between actions and outcomes are misspecified" (Levitt & March, 1988: 325). The claim of this dominant approach is that certain fashion setters use glitzy rhetorics to create the subjective experience in fashion followers that they are learning about rational and progressive management techniques. Fashion setters do so by exploiting fashion followers' sociopsychological vulnerabilities, not by providing This content downloaded from 149. 171. 67. 64 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 275 them with techniques that assist them in efficiently narrowing new and important performance gaps. I presented an alternative conception of fashion demand in which I have suggested that it is shaped by a competition between both sociopsychological and technoeconomic forces. With respect to technoeconomic forces, I argued that intraorganizational contradictions, as well as changes in their economic and political environments widen certain organizational performance gaps creating incipient preferences influencing management fashion demand.These gaps are more likely to shape management fashion if (a) fashion setters bring these gaps to fashion followers' collective attention and (b) make it possible for followers to learn about collectively acceptable management techniques capable of efficiently narrowing these gaps. Such a process more closely resembles real, as opposed to superstitious collective, learning (Miner & Haunschild, 1994). I also argued that business school scholars should not only study the management-fashion-setting process, but also use such studies as a guide for intervening in this process.The plea, then, is not to passively watch sociopsychological forces shape technically inferior management fashions, but to act in a scientifically informed

manner in order to render management fashion setting a more real, as opposed to superstitious, learning process, which adjusts organizations to changing organizational, political, and economic environments. The remainder of this section examines first what data might be needed to develop an understanding of management fashion. Then, the article's theory of management fashion serves as a framework to review a variety of prescriptions advanced by management scholars for anagement scholars interested in influencing management practice. Studying Management Fashion Figure 3 suggested three types of questions that need to be studied. The top part of Figure 3 raises the question: What is the relation between country norms and the existence and nature of management fashion markets? The middle part of Figure 3 raises the question: What is the structure of management-fashion-setting communities and how do managementfashion-setting processes function?

The bottom part of Figure 3 raises the question: What forces external to the fashion-setting process shape management fashion demand?One type of data may be particularly useful in addressing all three questions. I describe these data first and then suggest how management researchers could use them to address these questions. Fashion setters articulate rhetorics and disseminate them using popular and academic press articles. Such articles constitute, therefore, a large archival database useful in studying management fashion. These data cover long time periods, often are carefully indexed, and often are available in computerreadable formats. This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 76 Academy of Management Review January The relation between country norms and the existence and nature of management fashion markets could be examined

by studying articles from countries that differ with respect to their norms of rationality and progress (Guillen, 1994). Articles from various countries also could be employed to study the dissemination of management fashions across country borders (Kogut, 1991). The structure and composition of management-fashion-setting communities and how management-fashion-setting processes function also could be studied using articles that have a bearing on management techniques.Because each type of fashion setter has to disseminate rhetoric through a publication outlet, cataloging these outlets makes it possible to map out a fashion-setting community. Moreover, content analyses of titles, abstracts, or bodies of articles make it possible to study the content of fashionable management rhetoric (see e. g. , Guillen, 1994; Litterer, 1959). Authors' backgrounds, as well as where they publish articles, reveal what type of fashion setter disseminated what rhetoric when, making it possible to study the fashion-setting process (see, e. g. , Barley et al. , 1988).

Article counts, because they provide time-series data, may prove particularly useful when researchers study how forces external to the fashion-setting process shape management fashion supply and demand. I was able, for example, to use a time-series technique to analyze the yearly number of articles about various management techniques listed in the Business Periodicals Index between 1913 and 1993 in order to test hypotheses bearing on economic and political determinants of fashion demand (Abrahamson, In press). Of course, management articles are not the only way to study management fashion.Adoption data have been used to study the diffusion of fashionable management techniques (see, e. g. , Burns & Wholey, 1993; Fligstein, 1985)as well as whether fashionable management techniques are survival enhancing for organizations (see, e.

g. , Powell, 1995). Organizational and interorganizational level qualitative studies also provide insights about sociopsychological motives underlying the adoption and utilization of fashionable management techniques, as well as the creation and selection stages of the management-fashion-setting process (see, e. . , Covaleski & Dirsmith, 1988;Eisenhardt & Brown, In press; Ritti & Silver, 1986;Zbaraki, 1994). Influencing Management Fashion I have urged management scholars to intervene in the fashion-setting process and to be guided by an empirically validated theory of this process. I would be remiss, therefore, if I did not review prescriptions advanced by management scholars for management scholars interested in improving management practice. My goal, however, is not to comprehensively and critically review this prescriptive literature.It is rather to illustrate how my theory of management fashion can serve to classify This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 277 different types of prescriptions and reveal their complementarities. Which of these prescriptions are useful in shaping management fashion is a matter for empirical analysis guided by an empirically validated theory of management fashion.

Shaping norms of rationality and progress.If norms of rationality and progress open up markets for management fashions, then these norms might constitute one point at which management scholars could intervene to shape management fashion setting. Norms of rationality, for example, tend to call for one-most-rational-ways to manage that are often differentially efficient across contexts (Meyer & Rowan, 1977; Scott & Meyer, 1994). Therefore, Starbuck and Nystrom (1981) urged management scholars to promote norms that render organizational stakeholders tolerant of greater experimentation and diversity in fashionable management techniques.Others

have criticized the U. S. 's strong norms of progress and urged management scholars to promote a more balanced conception of management innovation as potentially retrogressive as well as progressive (Abrahamson, 1991; Eccles ; Nohria, 1992; Kimberly, 1981; Rogers, 1983). Such a more balanced conception presumably would lead to fewer transitory management fashions. Influencing the creation of management fashions. The creation stage involves the invention of management innovations and the rediscovery/ reinvention of old management techniques.Creation constitutes another intervention point for management scholars seeking to influence management fashion setting. One prescription is for scholars to make managers experiment, to create managerial innovations, or to seek out such natural experiments, in order to carry out rigorous qualitative studies of their consequences (Bettis, 1991; Nystrom ; Starbuck, 1977; Starbuck ; Nystrom, 1981). Another prescription is for scholars to study outliers (best and worst practices) rather than only averages, because outliers might reveal powerful management innovations (Bettis, 1991; Draft ; Lewin, 1990; Starbuck, 1993). Such tudies would foment greater awareness of technically functional and dysfunctional management innovations arising from natural experiments carried out by managers in response to shifting technoeconomic environments.

These studies also might cause management fashion setters to select the management techniques they attempt to launch into fashion from a more technically functional array of management techniques. Enhancing the selection of management fashions. Interventions by management scholars in the selection of management fashions could render management fashion setting a more technically functional learning process for fashion followers.There are a number of prescriptions suggesting how management scholars could alert fashion followers to new and important performance gaps and to management innovations that efficiently narrow these gaps. One recurrent

prescription has been for management scholars to be more timely (Beyer ; Trice, 1982). Both Bettis (1991) and Astley and Zammuto (1992: 456) urged management scholars This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 278Academy of Management Review January to be more aware and concerned with current developments in business so as not to lag far "behind the incidence of the phenomena it explains. " Others have suggested that the gaps and techniques that are selected for study should be decided upon with a greater concern for the preferences of fashion followers (Corwin ; Louis, 1982), with some prescribing and involvement of users of applied knowledge in the selection of areas requiring more scientific knowledge production (Lindblom; Cohen, 1979).Finally, still others have suggested that techniques and performance gaps should be selected for study in order to avoid falling well behind management fashion. Dean and Bowen (1994: 393), for instance, prescribed a focus on total quality (TQ) techniques in the 1994 AMR TQ issue because "given its importance in practice, we risk losing our credibility as management theorists by ignoring TQ in our research. " Modifying the processing of fashionable rhetoric.

Processing involves the elaboration of rhetoric that reveals organizational performance gaps and champions management techniques capable of narrowing these gaps.Management scholars have advanced a number of prescriptions, bearing on both the form and substance of scholarly rhetorics, which could render them more influential in shaping management fashion. With respect to form, management scholars have been urged to write clearly and forcefully, unencumbered by scholarly jargon (Beyer ; Trice, 1982). They have

been counseled to use qualitative data to make interpretations of quantitative data results more comprehensible, relevant, and convincing to themselves and to users (Beyer ; Trice, 1982).They also have been enjoined to study and write about constructs and techniques that can be manipulated by managers, revealing their techniques' performance implications, at not so abstract a level that managers cannot comprehend how abstract generalizations apply to their specific situations (Cheng & McKinley, 1983). Finally, they have been urged to include prescriptive implications as a normal part of disclosing research results (Bettis, 1991; Nystrom & Starbuck, 1977;Starbuck & Nystrom, 1981).Prescriptions bearing on the substance of scholarly rhetoric differ, depending on whether the prescriber advocates a more engaged or less engaged view of applied research. The most disengaged approach to applied science provides a total replacement of management knowledge with scientific knowledge. The prescriptions for the organization scientists are to develop a strong, unitary paradigm in order to speed up scientific understanding of managerial behavior (Pfeffer, 1993).The prescriptions for researchers are to contribute to normal science in order to develop a clear basis of knowledge on which scientifically tested prescriptions for change can then be offered to managers (Donaldson, 1992). Disengaged applied scientists preach separation from practitioners because management knowledge and priorities could bias and corrupt the substance of applied scientists' rhetorics (Dubin, 1976). Proponents of the more engaged approach to applied science see it as challenging and complementing managerial knowledge (Lindblom; Co-This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 279 hen, 1979). The prescriptions for the organizational sciences are to

maintain a rich diversity of perspectives and methods (Canella ; Paetzold, 1994).

The prescriptions for business school scholars are to use scientific theory and research to challenge and complement the assumptions guiding what managers think and do (Astley ; Zammuto, 1992; Davis, 1971).Engaged science advocates interaction with practitioners because helping them transform the world might help applied scientists understand it validated theory of (Starbuck ; Nystrom, 1981). An empirically management fashion might help determine whether engaged or disengaged approaches to management science are more efficient and effective in shaping management fashions and practices. Enhancing the dissemination of fashionable rhetorics. A broad variety of prescriptions also suggests how to reform the means by which management scholars disseminate their rhetorics so as to enhance management fashion setting.Prescriptions range from sponsoring junior faculty to help them establish contacts with management practitioners (Beyer ; Trice, 1982) to a radical restructuring of business schools in order to facilitate the dissemination of scholarly rhetorics. Beyer (1992) suggested creating formal positions in business schools for management scholars capable of (a) carrying out studies to exploit insights drawn from more basic applied research or (b) translating scholarly rhetorics into rhetorics fit for public consumption.Others advocate sponsoring joint conferences with practitioners, strengthening journals such as the Academy of Management Executive, which is directed at practitioners, using public relations firms to place reports of scholarly research in massmedia outlets, and forming alliances with mass-media editors and other management fashion setters (Hambrick, 1994). Still others advocate more direct means of reaching practitioners. Beyer and Trice (1982: 616), for example, urged management scholars to "act as their own self-advocates and disseminate their research findings in magazines read by

users as well as in professional journals. Improving the demand for management fashions. My theory of manfollowers are conagement fashion suggests that management-fashion sumers who influence the supply of management fashions. Training better consumers of scholarly management rhetorics constitutes, therefore, an important venue for intervening in the management-fashion-setting process. One prescription is to train undergraduate and MBA students to become better consumers of scholarly rhetorics (Beyer & Trice, 1982).Another suggestion is to modify the training of particular consumers of and human resource spescholarly management rhetorics-personnel cialists, research (inside) change agents, and consultant (outside) change agents (Beyer, 1992).

In conclusion, I have argued that management fashions are not cosmetic and trivial. Management fashions shape the management techniques that thousands of managers look to in order to cope with extremely important and complex managerial problems and challenges. One need This content downloaded from 149. 171. 67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 80 Academy of Management Review January only look to either the hundreds of firms and government institutions that are currently "reengineering" or the thousands of "downsized" employees that they are laying off, in order to realize that the management techniques that become fashionable have massive, sometimes helpful, but sometimes devastating, effects on large numbers of organizations and their employees. It follows from these observations that the study of management fashion is a very serious matter indeed, strongly deserving of careful inquiry by management scholars.I have sought to advance a theory of management fashion: a theory that could explain why and by what processes particular management techniques become fashionable and when they do. In this final section,

the article pointed to data and methods useful in studying management fashion. My only goal, however, was not simply to advance a theory useful in guiding empirical research on management fashion. Another goal was to develop a theory of management fashion that would reveal the position of business schools, and the scholars staffing them, in the managementfashion-setting process.This theory, if it were empirically validated, could allow management scholars to ascertain the utility of the various prescriptions reviewed in this final section, prescriptions advanced by management scholars, for management scholars interested in influenctheory, therefore, that could guide manageing management practice-a ment scholars intent on shaping management fashion and practice, not just studying it.

References

1. ABIInform. 1986. Ann Arbor, MI:University of Michigan, Inc. Abrahamson, E. 1991.
2. Managerial fads and fashion: The diffusion and rejection of innovations.Academy of Management Review, 16: 586-612. Abrahamson, E. In press.
3. The emergence and prevalence of employee-management rhetoric: Long-runand short-rundeterminants, 1913-1992. Academy of Management Journal. Determinants and consequences. AcadAbrahamson, E. , &Fombrun,C. 1994.
4. Macroculture: emy of Management of Management Review, 19: 728-755. Abrahamson, E. , &Rosenkopf, L. 1993.
5. Institutional and competitive bandwagons. Academy Review, 18: 487-517. Astley, G. , & Zammuto, R. 1992. Organization science, managers, and language games. Organization Science, 3: 443-459. Barley, S.R. , & Kunda, G. 1992.
6. Design and devotion: Surges of rational and normative ideologies of control in managerial discourse. Administrative Science Quarterly, 37: 363-399. Barley, S. R. , Meyer, G. W. , & Gash, D. C. 1988.
7. Culture of cultures: Academics, practitioners and the pragmatics of normative control. Administrative Science Quarterly, 33: 24-60. Barthes, R. 1983.
8. The fashion system. New York:Hill & Wans. This content downloaded from 149. 171.
   

67. 164 on Sat, 24 Aug 2013 23:34:16 PM All use subject to JSTOR Terms and Conditions 1996 Abrahamson 281 Bergler, E. 1953.

performance in scientific settings.Administrative Science Quarterly, 28: 85- 100. Corwin, R. G. , & Louis, K. S. 1982.