UNDERSTANDING STRATEGIC CHANGE: THE CONTRIBUTION …bigideasproject.wikispaces.com/file/view/greenwood1993.pdf · ® Academy of Managemenl JournoJ 1993, Vol. 36, No. 5, 1052-1081

® Academy of Managemenl JournoJ1993, Vol. 36, No. 5, 1052-1081.

UNDERSTANDING STRATEGIC CHANGE: THECONTRIBUTION OF ARCHETYPES

ROYSTON GREENWOODC. R. HININGS

University of Alberta

We examined the concept of archetype, implicit in a number of con-temporary approaches to the study of organizational design andchange. Despite an emerging interest in archetypes, the concept hasreceived inadequate investigation. The present article offers a hasis fordefinition of the concept and sets down three assumptions, which wetested using data collected longitudinally from 24 organizations. Wepresent a numher of theoretical and methodological implications of thearchetype approach to the study of organizational change.

The efforts of organizational theorists to understand the relevance ofalternative organizational designs and the dynamics of their evolution maybe entering a new phase. There appears to be a growing attraction to theuncovering of organizational archetypes, gestalts, or configurations andgrowing attention to their transformation and development (e.g., Johnson,1987; Miller & Friesen, 1980, 1984; Mintzberg, 1983).

An archetype is defined here in terms of two general statements. First,organizational structures and management systems are best understood byanalysis of overall patterns rather than by analysis of narrowly drawn sets oforganizational properties. This is the "holistic" perspective asserted byMiller and Friesen (1984). Second, patterns are a function of the ideas, be-liefs, and values—the components of an "interpretive scheme" (Ranson,Hinings, & Greenwood, 1980)—that underpin and are embodied in organ-izational structures and systems. An archetype is thus a set of structures andsystems that reflects a single interpretive scheme.

Interest in archetypes has proceeded alongside the study of organiza-tional change. Indeed, the central argument of the present research is thatarchetypes are important for understanding change. Researchers investigat-ing change have increasingly distinguished between incremental change(Lindblom, 1959; Quinn, 1982), and "frame-breaking" (Nadler & Tushman,1989), or "quantum" change (Hinings & Greenwood, 1988a; Miller & Friesen,1984; Tushman & Romanelli, 1985). Historically, interest in change has beenat a micro level (e.g., Bennis, Benne, & Chin, 1985), a focus that has producedan emphasis on incremental change within archetypes. The dynamics of thatprocess, however, are quite different from those of large-scale change, whichinvolves movement from one archetype to another.

An example of recent interest in frame-breaking change is the punctu-

1052

1993 Greenwood and Hinings 1053

ated-equilibrium thesis (Abernathy & Clark, 1985; Tushman & Anderson,1988; Tushman & Romanelli, 1985), whose roots are the life-cycle literature(Kimberly & Quinn, 1984). Organizations are described as experiencing pe-riods of convergence and stability punctuated by reorientations of strategyand design. Neither Abernathy and Clark nor Tusbman and colleagues usedthe term archetype, but their notions of "convergence" and "upheaval" im-ply both the idea of archetypes, as patterns of structural arrangements, anda definition of change as movement between archetypes.

Similarly, the population ecology school (Astley, 1985; Betton & Dass,1985; Carroll, 1984; Hannan & Freeman, 1977; Singh, House, & Tucker, 1986)shows attention to archetypes. These researchers define populations of or-ganizations in terms of organizational "forms" having common vulnerabilityto particular environmental circumstances. The term form is impreciselydefined but is similar to archetype in that it covers the configuration of workpatterns. The population ecologists' empirical concern is actually withchange in the population of archetypes—their emergence and disappear-ance—but it is the underlying concept of archetype that is important as apreliminary to our argument.

Despite its contemporary relevance to theorizing about change, the ideaof archetypes is inadequately treated in the literature, possibly because the-orists are often concerned only secondarily with defining and establishingthe existence of archetypes. Much greater interest has been shown in theextent to which, and the conditions under which, organizations change, ormove between archetypes. As a consequence, there is no adequate definitionor basis for the uncovering and classification of archetypes. And yet, in orderto get to grips analytically and empirically with strategic organizational re-orientations, it is necessary to conceptualize the scale of change. The con-cept of archetype gives a basis for doing this by defining what constitutesframe-breaking, strategic change.

This study had three purposes. First, we sought to define the idea of anarchetype, highlighting the key role of the underlying interpretive scheme.In doing so, we clarified the main ideas and assumptions of the archetypeapproach in terms of a core assumption and three hypotheses. Second, weexamined the predictions empirically in order to assess whether the idea ofarchetypes warrants serious further consideration as a means of understand-ing organizational change. Third, we reviewed theoretical and methodolog-ical issues involved in the study of archetypes, as they are defined here. Inmany respects, this third purpose is reflexive and involves drawing anddistilling implications for future research.

ARCHETYPES

The development of interest in archetypes, necessitated by interest inorganizational change, represents a return to a central thrust of organization-al theory, which is the need to understand organizational diversity throughtypologies (Lammers, 1978; Weber, 1947). The concept of an archetype im-

1054 Academy of Management /ournai October

plies some form of classification. The idea of coherence between the ele-ments of organizational arrangements is central to typologizing, and theclassification of organizations is made according to differences and similar-ities in overall patterns. As Van de Ven and Drazin wrote, "Subcomponentsare related to each other in ways that yield a coherent ensemble (i.e., anoverall pattern called organization design). In organization theory these el-ements have been referred to as ideal types, modes, programs, populations,etc. Much of our theorizing is explicitly in terms of types . . . [that is] basicpatterns of organizing that are coherently designed to yield a systematicconfiguration to the components" (1985: 348). Uncovering patterns of orga-nizing, however, involves two tasks: first, definition of the elements of or-ganizational design; second, resolution of how to recognize and interpret apattern of those elements.

Elements of Organizational Design

The search for Van de Ven and Drazin's "basic patterns of organizing"requires a multidimensional definition of design. Typically, the emphasis inorganizational classification has been structural, with analysts concernedwith the differentiation of tasks and positions, the formulation of rules andprocedures, and the prescriptions of authority (cf. McKelvey, 1975; Miller,1981; Mintzberg, 1979; Pugh, Hickson, & Hinings, 1969). This approach isimportant because of its emphasis on structural forms as instruments forachieving calculable and predictable control of organizational performance.However, it has resulted in structure being taken as definitive of organiza-tional design. Essentially missing from such definitions are the processesand systems that connect and activate structural frameworks. As Pinder andMoore (1979) suggested, multiple parameters need to be employed to cap-ture the complexity of organizational similarities and differences.

Following Daft and Macintosh (1984), we would suggest that an accept-able definition of organizational design has to include coverage of bothstructure and systems and that there is a developing convergence upon adefinition of design as having three essential elements: (1) the vertical andhorizontal structures of roles and responsibilities, (2) decision systems, orpolicy and resource allocation mechanisms, and (3) human resource sys-tems, such as recruitment, appraisal, and compensation. Each element hasbeen analyzed and refined in previous work. The recurring concern is withhow organizations manage the complementary issues of differentiation andintegration involved in vertical and horizontal allocations of authorities andresponsibilities. Commonly used structural dimensions (for review, seeChampion [1975], Khandwalla [1977], and Mintzberg [1979]) are the extentof differentiation (e.g., Blau & Schoenherr, 1971); the criteria of differentia-tion, or the basis for grouping activities (e.g., Blau & Schoenherr, 1971); andthe extent of integration (e.g., Galbraith, 1975; Lawrence & Lorsch, 1967).

Miller's (1987) survey identified three dimensions of decision systems:rationality, interaction patterns, and assertiveness. Rationality is the extentto which decisions and resource allocations are the product of careful anal-

ig93 Greenwood and Hinings 1055

ysis, systematic scanning of markets for problems and opportunities, andmethodical planning and articulation of strategies (Andrews, 1980; Porter,1980; Quinn, 1982). Interaction is the degree to which decision making isindividual and directive or collective and consensual (Pettigrew, 1973; Wil-davsky, 1984). Assertiveness covers the extent to which an organizationengages in reactive or proactive behavior and is prepared to assume risk inpursuit of competitive advantage (Miles & Snow, 1978).

Human-resource-system dimensions of organizational design are lesswell developed than are the structural dimensions, despite growing appre-ciation of the former's importance (Cummings, 1984; Dyer, 1983; Stonich,1984). Significant dimensions of human resource systems are the extent towhich control is hased on a "hureaucratic" or a "clan" form of appraisal(Ouchi, 1977) and the criteria used for recruitment, career development andcompensation—whether professional or managerial competence is domi-nant (Cummings, 1984; Dyer, 1983; Schuler, 1984; Stonich, 1984).

The Basis of Pattern and Classification

Delineating elements of organizational design enables us to deal withour original statement that structures and management systems are hest un-derstood in terms of overall patterns rather than by analysis of narrowlydrawn sets of organizational properties. But a central problem within thearchetype thesis is how to uncover coherent patterns of organizing. Millerand Friesen (1984) used statistical manipulation of large samples of organ-izations, observed over periods of 20 or more years, to produce configura-tions of relationships between variables. Mintzberg (1979,1983) constructedideal types through a combination of analytical deduction and an extensive,though selective, synthesis of existing research. For us, "The starting pointfor identification and classification of design archetypes is . . . identificationof the interpretive scheme and of how that relates to structural attributes andprocesses" (Creenwood & Hinings, 1988: 299).

The pattern of an organizational design is a function of an underlyinginterpretive scheme, or set of beliefs and values, that is embodied in anorganization's structures and systems. An archetype is thus a set of struc-tures and systems that consistently embodies a single interpretive scheme.Defining an archetype in this way is a departure from the more commontreatment of structures and systems as disembodied attributes of organiza-tions related in an adaptive way to context and performance. Structures andsystems, from our perspective, are not neutral instruments but embody—wittingly or otherwise—intentions, aspirations, and purposes (Barley, 1986;Bartunek, 1984; Clegg, 1975; Perrow, 1982).

A range of writings reflects this view of archetypes as configured struc-tures expressing underlying values. Ranson, Hinings, and Creenwood main-tained that to fully understand the constitutive structuring of organizationalstructures over time, an investigation of the "social mechanisms that deter-mine the process of structuring and shape the ensuing structural forms" isnecessary (1980: 3). Miller, although he did not explicitly include values as

1056 Academy of Management Journal October

part of the generation of configuration, wrote of "central, enduring themesthat unify and organize" (1987: 697), an idea taken further in The IcarusParadox (Miller, 1991). Tushman and Romanelli (1985) emphasized the im-portance of the development of new values, norms, beliefs, and rationaliza-tions supporting new strategies, structures, and systems. Their concept ofconvergence implies that strategies and structures already in place are for-mulations of values that provide both a logic and a propellant for an organ-ization.

In a study of architectural practices, Blau and McKinley (1979) elabo-rated a variety of "work motifs," each of which entails a distinct relationshipto basic dimensions of organization. A work motif is an intellectual ethos, aset of ideas about the practice of architecture, that results in a particular setof organizational arrangements. This idea is similar to Karpik's (1978) con-ception of "logics of action," or modes of rationality within organizationsthat designate "forms of coherence" (cf. Callon & VignoUe, 1976).

In his work on ICI, Pettigrew (1985) wrote of "dominating rationalitiesor core beliefs" articulated through structural design and organizational pro-cesses. Thompson (1975), discussing religious organizations, elaborated theconcept of "symbolic appropriateness," whereby organizational practicesare geared to the theological position adopted by a church or sect.

These approaches suggest that sets of ideas that have the status of val-ues—or preferences for certain courses of action and outcomes (Beyer,1981)—underpin organizational structures and processes. We are positingthat archetypal coherence comes from the consistent relationship betweenan interpretive scheme and an organization's structures and systems.

There are two convincing theoretical reasons for anticipating that organ-izations will develop structures and systems consistent with a single inter-pretive scheme. Miller and Friesen (1984) provided one such explanationwith their concept of "momentum," describing organizations as evolvingtoward archetypal coherence because for any firm it is better to be one thingconsistently than to be a combination of ill-fitting parts. In effect, they ac-knowledged the economic benefits that flow from coherence. A rather dif-ferent explanation recognizes organizations as composed of groups whosepositions of relative advantage and disadvantage are shaped by the organ-ization's design Qohnson, 1987; Pfeffer, 1981; Walsh, Hinings, Greenwood, &Ransom, 1981). Structures and systems allocate scarce and valued resourcesand indirectly legitimate and perpetuate distributive inequalities by the con-sistency of the cues and messages transmitted. An organization's "dominantcoalition" will seek to remove discordant structures because of the risk ofchallenge to the legitimacy of the status quo. Organizations will thus evolvetoward archetypal coherence as advantaged groups seek consolidation ofpolitical position and control over the distribution of resources.

Critically, the position developed here departs from earlier work onculture, a term we do not use, which also embraces values, beliefs andorganizational practices. For our purposes, it is essential to distinguish be-tween the values and beliefs—the interpretive scheme—embodied in an


organization's structures and systems and the extent to which some or allorganizational actors are, at a given time, committed to those ideas andvalues. The term culture is often used to cover both the values and beliefsexpressed through structures and systems and the degree to which actorsaccept and behave in accordance with those values and beliefs. Researchershave seen culture as a system of shared beliefs that underpins structures,commitments, and actions (Kilmann, Saxton, & Serpa, 1985; Sathe, 1985;Schein, 1986). A strong culture thus usually refers to circumstances inwhich core values and beliefs are widely held throughout an organizationand practices and procedures are consistent with those values and beliefs. Incontrast, for us the pattern of commitments to one or more interpretiveschemes is a potential dynamic of change. For example, an organization'sprevailing interpretive scheme (the one embodied in structures and systems)need not be supported by any actors in the organization. At some point in theorganization's history it would have been supported, but over time thatcommitment, for various reasons, waned. However, the structures and sys-tems have remained unchanged because of organizations' known tendencytoward inertia. Such a situation would be relatively fragile and amenable tochange. We offer these comments to highlight the importance of viewing apattern of commitments as a dynamic, not a definitional, component ofchange.

Archetypes and Institutional Specificity

The discussion thus far has defined an archetype as a set of structuresand systems consistently reflexive of a single, underpinning interpretivescheme. An important implication of this definition, with its emphasis uponthe core significance of the interpretive scheme, is that archetypes are prob-ably institutionally specific.

Organizational theory has traditionally sought to produce theory gener-alizable across broad classes of organizations. Recently, however, an increas-ing body of knowledge and research has pushed in the direction of whatChild and Smith (1987) referred to as sectoral analysis (e.g.. Huff, 1982;Pettigrew, 1985; Spender, 1980; Worthy, 1985). There is disillusionmentwith overly abstract studies that purport to embrace all organizations, mod-eling tbem as disconnected from the contexts that provide meaning to tbeiractivities. Institutional theory clearly points to the importance of ideas andvalues within a sector that limit the range of likely organizational designs. Ina review of institutional tbeory, Scott wrote, "Institutionalized belief sys-tems constitute a distinctive class of elements that can account for tbe ex-istence and/or elaboration of organizational structure" (1987: 497). In otherwords, values and beliefs external to an organization play a significant rolein determining the manner in which it operates. Meyer and Rowan (1977)originally developed this argument by suggesting tbat organizations conformto sectoral rules and requirements for reasons of legitimacy and resourceflows, an argument that DiMaggio and Powell (1983) and Tolbert and Zucker

1058 Academy of Management /ournaJ October

(1983) elaborated. More recently, Baum and Oliver (1991) showed that or-ganizations increase their survival capabilities by conforming to institu-tional norms.

This brief review suggests that the search for archetypes should startwithin an institutional sphere or industrial sector, for it is there that insti-tutional prescriptions and proscriptions (Hinings & Greenwood, 1988b), orwhat Child and Smith (1987) and Johnson (1987) referred to as "strategicrecipes," will be found.

HYPOTHESES

Two initial hypotheses can be derived from the discussion, both dealingwith the existence of archetypes.

Hypothesis 1: Organizations tend to operate throughstructures and systems that are manifestations of a single,underlying interpretive scheme: they exhibit archetypalcoherence.

Hypothesis 2: Organizations that have structures and sys-tems that are not manifestations of a single, underlyinginterpretive scheme iviil move toward archetypal coher-ence.

The archetype thesis suggests that change involving movement betweenarchetypes is highly unusual. Organizations are characterized by conver-gence toward a prevailing archetypal form and inertia and tend to remainwithin the assumptions of the existing archetype (Tushman & Romanelli,1985). "Conservatism" (Child and Smith, 1987) marks normal organizationaldevelopment, with change "configured in terms of the [existing] paradigm"Qohnson, 1987: 245). Radical change, the passage from one archetype toanother—"frame bending," to use Nadler's (1987) phrase—is exceptionalThus,

Hypothesis 3; Organizations tend to remain within an ar-chetype rather than move between archetypes.

It is worth stressing, however, that the hypotheses are based on thecentral assumption that archetypes that are discoverable by researchers doexist in institutional arenas and that actual organizational arrangements ap-proximate those archetypes to a greater or lesser degree. In other words,institutionally located prescriptions and proscriptions constrain and influ-ence the design of organizational arrangements, in a fashion unrecognizedby organizational participants. In approaching the hypotheses, therefore, ourinitial task was to validate the assumption of institutional prescriptions byexposing the archetypes operant within institutional settings.

A distinction is being drawn between the hypothesized existence ofinstitutionally prescribed archetypes and their hypothesized, empiricallyassessable appearance in individual organizations within an institutionalsector. Empirically, it is known that organizations exhibit complexity and


variety when examined along dimensions such as the structure of roles andresponsibilities, decision systems, and human resource systems. As a result,organizational typologies and taxonomies tend to produce multiple forms(McKelvey, 1982; Mintzberg, 1979; Pugh et al., 1969). Similarly, when onelooks for the empirical existence of archetypes, variety will be found, as ourdata show. We are thus not saying that organizations do not vary becausearchetypes exist. We are saying that within a given institutional sphere, onlya small number of archetypes are legitimated. The number will vary with theextent to which a particular sector is institutionalized. In a very heavilyinstitutionalized sector, there may be only one or two institutionally ap-proved forms at a given time (Hinings & Greenwood, 1988a; Oliver, 1991;Kikulis, Slack, & Hinings, 1992).

Our point, in short, is that two levels of analysis are pertinent. The firstis the institutional sector, where the research task is to discover which organ-izational forms are legitimated. These legitimated forms are the archetypetemplates and will be few in number. The second level is that of individualorganizations, where the research task is to examine the extent to whichthose organizations do or do not approximate their sector's archetype orarchetypes. At this level, there will be more variety, but pressures for arche-typal conformity will operate upon individual organizations.

METHODS

In describing the methods we used, we are aware that certain method-ological features are important implications of the archetype thesis itself. Soa later section reflects not only on the theoretical implications of the study,but also on important methodological implications of the theory as theyemerged from the research. In this sense, we depart from the more conven-tional paradigm of discussing methodology as if it were an abstract anddefinitive set of procedures. Instead, we described our methods in a mannerthat permits their subsequent discussion.

Following Van de Ven and Drazin's suggestions (1985: 348), we used atwo-stage research design. In the first stage, we examined the assumptionthat ideal archetypes exist intellectually within a given institutional sector;this stage involved the identification of archetypes for use as templatesagainst which the organizational arrangements of a sample of organizationscould be compared. In the second stage, we examined whether the organ-izational arrangements of the organizations corresponded to the templates.Table 1 summarizes the methods used in each stage.

Stage One: Identification of Archetypes, 1969-72

To establish an archetype requires two steps. First, the underlying in-terpretive schemes within an institutional sector have to be isolated. Second,the structural and processual implications of these interpretive schemeshave to be uncovered. As we defined them, archetypes are what McKinneycalled a "constructed type," "a purposive, planned selection, abstraction.

1060 ^ Academy of Management Journal October

TABLE 1Research Strategy Summarized

Stage Methods1. 1969-72

Identification of interpretive schemes • Documentary analysis of the sectorwithin the institutional sector • Unstructured interviews, 200 respondents

1971-72Documentary analysis of population to • Structured questionnaire, 72 percentproduce sample of 24 organizations response from population of 422, and

documentary analysis of specificorganizations

• Analysis of academic research on localgovernment

2. 1974-82Longitudinal examination of 24 • 752 semistructured interviewsorganizations • Documentary analysis

• Structured questionnaire with 100-percentresponse

combination, and (sometimes) accentuation of a set of criteria with empiricalreferents that serves as a basis for comparison of empirical cases" (1966: 3).

The archetypal templates hunted in the present research were distilledfrom the prescriptions operating within the intellectual milieu of a singleinstitutional context. Sector specificity was taken as a given. Admittedly,having data from several sectors would have strengthened the findings re-ported in the second stage of the research; however, the difficulties involvedin unearthing interpretive schemes, combined with the demands of a longi-tudinal methodology, provided a strong pragmatic justification for restrict-ing attention to a single sector.

The particular institutional sector examined was municipal governmentin England and Wales. These organizations, called "local authorities," differfrom their counterparts elsewhere because of their size—they employ up to50,000 people—and functional complexity. The range of functions per-formed includes education, social welfare, highway construction, consumerprotection, and recreational and cultural services. These are large, complexorganizations.

The municipal sector was chosen as the research milieu because it metseveral criteria. First, the sector is well-defined and the organizations withinit relatively homogeneous in terms of their legal status and the complexity oftheir responsibilities, thus ensuring that comparisons would be of like withlike. Second, the organizations had discretion to organize as they wished,thus avoiding the problem of constraint by parent companies (cf. Child,1973; Johnson, 1987). Third, the system of local government in England andWales was recast on April 1, 1972, and each of the 422 organizations exam-ined here had to produce reorganization plans (Dearlove, 1979) that were to


take effect in 1974. This process led to considerable debate, which raisedawareness of underpinning ideas, making them amenable to recognition.Fourth, access was nonproblematical, an important consideration when re-search demands upon participating organizations are to be extensive for alengthy period (Doz & Prahalad, 1988).

The search for archetype templates had three concurrent and iterativesteps:

Context documentary analysis. Three researchers, including the presentauthors, carried out content analysis of reports and documents dating fromthe late 1950s to the late 1960s, pertaining to the organization and manage-ment of municipal authorities, and written by representatives of the Britishcentral government, local organizations, professional associations, and uni-versities.^ Each researcher identified statements and normative referencesthat dealt with the appropriate domain of the municipality, how it shouldorganize and manage itself, and performance expectations. Typical state-ments concerned such issues as failures to plan, the absence of clear goalsand objectives, professional insularity, the lack of coordination betweenprofessions, and the need for new structures, such as a policy-planningprocess and a chief executive. The researchers met regularly and clarifiedthe central themes of the documents.

Interviews and project groups. From 1969 to 1971, the researchers tookadvantage of residential ten-week executive development courses held at theUniversity of Birmingham. We interviewed 200 senior managers from ap-proximately 100 organizations using open-ended interviews. Table 2 sum-marizes the interview guide employed. The interviewees were from 30 to 45years of age and had usually worked for at least three municipal organiza-tions. They held senior managerial responsibilities and were regarded bytheir employers as upwardly mobile. The interviews lasted from one to threehours, were conducted in informal settings, such as after meals, and ofteninvolved more than one researcher and two (sometimes three) interviewees.These interviews, though informal, were an extremely rich source of databecause they were conducted in an educational environment in which prac-ticing managers were stimulated to reflect upon, and debate, ideas aboutorganization and management in their institutional context in a manner notusually possible in one-on-one interviews in a work setting. Complementingthe interview program were observations of discussions among managers inproject groups of ten; the groups' formal task was designing an appropriateorganization for a post-reorganization municipality, and the group discus-sions totaled 12-15 hours over three-week periods. We recorded and dis-cussed the main themes and ideas of the discussions outside the groups.Each project group prepared written reports, which were presented orally tocolleagues and the researchers, who retained them for future reference.

' Greenwood and Stewart (1974) and Stewart (1971) summarize some of this material.

1062 Academy of Management Journal

TABLE 2Interview Guide

October

Section Illustrative Questions

Ideas and values

Structures and systems

What do you see as the proper role of a municipalorganization? What is its primary purpose? Is it changing? Ifso, what are the reasons and direction of change?

How is the nature of your job changing, if at all? What arethe purposes behind the change? How should it be changed?How might it change?

What do you see as the important changes required ofmunicipal organizational arrangements? What changesshould the municipality be attempting?

What are the key required changes?What are the key performance factors presently used in

your organization? Are they changing?What are the main changes occurring in municipal

structures and systems?What is the role of the chief executive officer and the

management team? How do they differ from the town clerkand the chief officers' group?

What is the reason behind the interest in corporateplanning and how does it differ from more conventionalbudgeting processes?

Organizational document analysis. Prior to the reorganization of localgovernments, the population of 422 municipal organizations was requestedto send copies of all reports and documents relating to organizational andpolicy-making arrangements that would be implemented after the reorgan-ization: 304 organizations (72%) responded. The materials provided not onlydescriptions of organizational structures but rich narrative accounts explain-ing the organizations' purposes and their managers' ideas about and expec-tations of the way the organizations would operate. We content-analyzed thedocuments in terms of the emerging themes, tbe tbree researchers jointlyreviewing each report.

The three-year program of documentary analysis and interviews, al-tbougb somewbat messy at the outset, gradually converged on our elabora-tion of two sets of sectoral values, or interpretive scbemes, and tbeir respec-tive implicated organizational arrangements. Throughout, we met frequentlyand discussed tbe growing collection of materials, comparing and examiningthem, often in an iterative fashion. By late 1970, two archetypes were rea-sonably salient in outline form: tbe beteronomous professional bureaucracy(Scott, 1965) and the corporate bureaucracy. Tbe features of tbese archetypesare described below and summarized in Table 3. The archetypes were notarticulated fully in the documentary sources in tbe coherent way set out inthe table. Tbe sources provided paper trails of incomplete ideas, cues, andexpressions tbat required final intellectual assembly.

Tbe arcbetypes were validated in two ways (Yin, 1984). First, we intro-

1993 Greenwood and Hinings

TABLE 3Components of Two Organizational Archetypes

1063

Components Corporate BureaucracyHeteronomous

Professional Bureaucracy

Interpretive schemeDomainPrinciples of

organizingSelf-evaluative

criteria

Organization designStructures

Criteria ofdifferentiation

Ratio of integrationto differentiation

Human resource systemsCriteria for reward

distribution

Type of controlDecision systems

Rational analysis

Interaction pattern

Assertiveness

Community governanceCentralized, corporate

integrationAnalytical appraisal by

interprofessionalmanagement

Program or clientele

High

Recruitment, careerdevelopment, andrewards based onmanagerial andprofessionalcompetence

Bureaucratic

Nonincremental, highlyrational

Based on corporate goalsand cross-professionalinteraction andinvolvement

Proactive stance towardpolicy making

Local administrationProfessional differentiation

Professional practice definedby intraprofessionaljudgment

Professional groups

Low

Recruitment, careerdevelopment, and rewardsbased on professionalqualifications andexperience

Clan

Incremental, less rational

Based on professionalinvolvement andjudgment; no corporateaccountability for goals

Service administration,essentially reactive

duced them as intellectual constructs to 30 academics and chief executiveswho were prominent members of the local government community. Second,the ideas were presented in the residential executive development coursesreferred to above. The purpose was to verify that we had properly under-stood the interpretive schemes and their organizational implications andthat the schemes were meaningful in the focal institutional sector. The levelof agreement was extremely high concerning the overall existence of the twoarchetypes; there were almost no dissenters. The exact organizational im-plications were subject to somewhat more discussion, but a strong concen-sus emerged quickly.

Thus, validation was obtained, and the existence of two competing ar-chetypes established.

The heteronomous professional bureaucracy archetype conceives organ-izations as administrative vehicles for the delivery of essentially disparate


services, such as education, social welfare, highway construction, housing,and police and fire protection. In this conception, the range of a municipalgovernment's responsibilities is a consequence of historical accident ratherthan of policy intent, and the domain of the local authority is the sum of itsresponsibilities, each treated as a separate activity. The corporate bureau-cracy archetype, in contrast, emphasizes the organization as an instrumentof community governance. Within this conception, local services are strandsof an integrated strategy for managing and influencing the organization'sgeographical community. The role of the organization is not administrative,implementing legally prescribed services, but governmental, combiningpackages of services and interventions in a strategic fashion.

These alternative beliefs about purpose and domain (community gover-nance versus local administration) are connected to ideas about how theorganization should evaluate its effectiveness in discharging its role. Thecorporate bureaucracy archetype emphasizes a high measure of analyticalappraisal combined with rigorous assessment of programmatic (interservice)performance. Clear statements of purpose are to be made, targets established,and performance monitored. Attention focuses upon the reciprocal effects ofservices upon client or focal groups. Appraisal of performance is the respon-sibility of senior management. In the heteronomous professional bureau-cracy archetype, on the other hand, professional autonomy and judgment arecritical. There is no conception of general management. The primary per-formance criterion is professional competence within each service area, andgood performance is the best practice as defined by professionals.

In other words, the structure of accountability in the corporate bureau-cracy transcends professional jurisdictions and imposes analytical rigor andclarity of performance targets, but the heteronomous professional bureau-cracy sets accountability within professional jurisdictions.

The two sets of values are each translated into structures and systems,thus constituting archetypes. Structurally, the heteronomous professionalbureaucracy organizes services as separate, highly specialized units basedon professional competencies. There is a low ratio of integration to differ-entiation, and the criterion of differentiation uses the profession as the basisof departmentation. Recruitment and career development systems empha-size the importance of professional qualification: only professionals can be-come department heads, and administrators are of low status. Compensationand appraisal systems focus upon performance of professional activitiesand, perhaps most important of all, the resource allocation mechanisms andthe underpinning information system are arranged in terms of Wildavsky's(1984) incremental, nonanalytical model of budgeting, a model suited to thevalues of the professional bureaucracy (Greenwood, 1984).

The corporate bureaucracy has a different organizational framework. Itemphasizes use of integrative devices, such as a chief executive officer,program directors, and central analytic capabilities imposed across servicedepartments. Hence, there is a high ratio of integration to differentiation.Programs, not professions, are the basis of departmentation: departments are


organized on the basis of objectives and outputs. Recruitment emphasizesgeneral management competence combined with professional experience.Incentive systems reward corporate rather than professional contributions.Career structures facilitate movements across professional boundaries, forinstance, by employees' assignment to central agencies or through theirmembership in interdisciplinary project teams. Resource allocation mecha-nisms encompass programmatic as well as professional review. Allocationsreflect corporate goals, subordinating professional to managerial judgmenton the basis of analytical techniques (Skelcher, 1980).

Stage Two: Empirical Examination, 1972-82

In the second stage of the research, we examined whether the organiza-tional arrangements of practicing organizations corresponded to the identi-fied templates in the manner set out in the three hypotheses.

Sampling procedure. A sample of 24 organizations was drawn from the304 organizations that provided documents. The documents outlined man-agement structures to be put in place in the post-reorganization period.Decisions on planning and human resource systems were not to be madeuntil after the reorganization had occurred. We used two measures of struc-ture, the ratio of integration to differentiation and the use of program orprofessional-functional criteria of differentiation, to construct a two-by-twomatrix, shown in Figure 1. One quadrant of Figure 1 describes the corporatebureaucracy (high ratio plus high use of program criteria), and another de-scribes the heteronomous professional bureaucracy (low ratio plus high useof professional-functional criteria). We drew organizations from each quad-rant, controlling for size by taking both large and small organizations fromeach quadrant (Hinings & Creenwood, 1988a).

Figure 1 also shows that 158 organizations were between archetypes atthe beginning of the research period. There are two reasons for this occur-rence. First, it is in the nature of archetypes, adaptation, and processes oforganizational change that there will always be organizations that are not ina complete state of fit. They may be close to archetype status or some dis-tance away ("embryonic" and "schizoid," Creenwood & Hinings, 1988).Second, in an institutional setting in which pressures for change are partic-ularly strong, significant numbers of organizations are likely to be in non-archetypal situations. This was the case in the setting studied here.

The sample size of 24 balanced our desire for a reasonably large sampleagainst our pragmatic ability to conduct detailed case studies of each organ-ization over three periods, 1974-75, 1977-78, and 1981-82. Because ouroverriding concerns were to examine the extent to which the newer arche-type, corporate bureaucracy, had actually been established and sustainedand to understand the dynamics of change, we used a purposive samplingframe: 6 organizations each were drawn from cell b and cell d, 9 from cell c,and 3 from cell a. Hinings and Creenwood (1988a) contains further details.

Data collection. The primary method of data collection was a program ofsemistructured interviews with the chief executive and senior managers (the


FIGURE 1Sampling Matrix

Criteria ofDifferentiation

Program

Professional-functional

Ratio of Integration to DifferentiationLow High

(a) Incoherent

107

(b) Goherent: heterono-mous professionalbureaucracy

29

(c) Coherent: corporatebureaucracy

117

(d) Incoherent

.51

department heads) of each organization; at least six managers were inter-viewed per organization. A total of 752 interviews were conducted, eachlasting approximately one and a half to two hours. Interviews concentratedon allowing respondents to explain their organization's purpose and tasks,how resources were allocated, how policies were set, and the interactionsbetween departments. We conducted the interviews in 1974-75 and re-peated them in 1977-78 and 1981-82. Surprisingly, there was virtually noturnover in the personnel interviewed, the likely explanation for that stabil-ity being that there had been many retirements and personnel changes in1972.

Measurements. We obtained the ratio of integration to differentiation bycalculating the number of lateral policy and budget committees in a munic-ipal authority as a percentage of the number of separate departments it had.A high value indicated a high integrative effort. The measure of the criteriaof differentiation was calculated as the absolute number of departmentsbased upon a program rather than a functional-service basis. A functional-service organization would have departments such as engineering, planning,and architecture, whereas a program organization would have a departmentof technical services combining the previously separate units into team-based units. A high value indicates high use of the program criterion. Theinteraction pattern was measured as the extent to which certain policy andmanagerial issues were routed through interdepartmental structures. Weidentified seven issues: preparation of draft estimates, finalization of reve-nue estimates, preparation of the capital investment program, recruitmentand promotion of senior managers, and three substantive policy issues rel-evant to each department analyzed; the pattern of accountability for eachissue was then mapped. The distinction sought was whether issues were


routed through a management team with multiprofessional membership orhandled exclusively within a given professional department. Tbe range ofpossible ratings for eacb issue was from 0 to 3, witb a bigb rating indicatinga corporate pattern of interaction. Ratings were based upon tbe number ofprofessions involved in discussing eacb issue. The organization's interactionscore was tbe total for tbe seven issues.

The degree of rational analysis was measured as the extent to which anorganization employed eight techniques, such as "position statements,""needs-gap analysis," and "output budgets," wbose use constituted part oftbe wider debate on the applicability of corporate planning and budgeting tomunicipal autborities (cf. Skeicher, 1980). Organizations were rated on athree-point scale (yes = 3, partly = 2, and no = 1) for each technique. Thehigher the rating, the greater tbe use of rational analysis within a resourceallocation system.

Tbe extent of assertiveness was measured as tbe extent to wbicb anorganization had explicit procedures for three items: overall environmentalscanning, an explicit meta-policymaking system covering the organizationas a whole, and explicit environmental scanning procedures witbin eachdepartment. Organizations were rated on each item in terms of the samethree-point scale described in the preceding paragraph.

The attempt to measure the criteria of reward distribution proved abor-tive. Tbe organizations were governed by national terms and conditions ofemployment; as a result, we observed little variation in rewards. Similarly,attempts to identify the basis of recruitment for senior positions were un-successful because of the sensitivity of tbe issue witbin the organizations.

The distinction between clan and bureaucratic forms of appraisai wasindexed as tbe extent to wbicb a department head's policies and programwere scrutinized using a common format of cross-organizational goals andpriorities rather than professional goals relevant to a single department. Wedefined bureaucratic control as occurring wben departmental policies andappraisals were subject to examination by corporate bodies such as manage-ment teams, corporate evaluation groups, and policy committees. Clan con-trol was seen as occurring when departments were allowed to operate ac-cording to their own internal standards. Table 4 gives basic statistics andalpha coefficients for all measures.

We used tbe interviews to complete a narrative description of eacborganization studied, wbicb was then validated in two ways (Yin, 1984).First, and of greater weight, documents, such as organizational charts, bud-getary procedures and reports, policy statements, and appraisal reviews,were used to confirm the narrative. For example, we validated the routing ofissues (the measure of interaction patterns) by examining a selection of com-mittee reports and agendas to see which committees and task forces lookedat specific reports. Similarly, whenever interviews led to the conclusion thatan organization was using techniques of rational analysis, we requestedcopies of reports illustrating tbeir utilization. We referred to documents notsimply to confirm the veracity of the respondents, but also to compare and

1068 Academy 0/ Management Journal October

classify organizations reliably. It is easier to compare organizations alongdimensions like rationality when operational examples (the use of specifictechniques) from different organizations can be directly examined and com-pared.

A second procedure used to validate the narrative account of each or-ganization's structures and systems was to have a senior officer in the organ-ization read it. We used this second validation procedure in approximatelyhalf of the sample. In every case, there was agreement with the account.

Organizational patterns. We computed standardized scores for eachvariahle using a mean of 50 and a standard deviation of 15. Standardizedscores were used to enable comparisons across variables and across timeperiods with regard to the extent to which organizations were in an arche-type. An organization was classified as within the corporate bureaucracyarchetype if it scored two-thirds of a standard deviation or more above themean on each variable, achieving a score of 60 or more. It was classified aswithin the heteronomous professional bureaucracy archetype if it scoredtwo-thirds of a standard deviation or more below the mean on each variable,achieving a score of 40 or less. We chose two-thirds of a standard deviationas the most appropriate cutoff point because a greater standard deviationwould have resulted in a large middle range including most scores, and alower standard deviation would have created a narrow middle range withmost scores falling in the high or low extremes. An organization was clas-sified as within an archetype if all its structures and systems were consistentwith the theoretical template.

TABLE 4Basic Statistics of Organizational Design Variables^

Variables

StructuresCriteria of

differentiationRatio of

integration todifferentiation

Human resourcesystems

Appraisal andcontrol

Decision systemsRational analysisAssertivenessInteraction

pattern

Time

Means

10.29

7.08

11.04

15.384.58

11.13

1

S.d.

3.37

4.30

3.11

7.881.98

2.17

Survey 1Points

Time 2

Means

10.29

7.38

14.58

22.215.21

11.75

s.d.

3.29

4.40

1.72

9.682.17

2.01

Time

Means

10.17

8.75

11.92

15.084.96

11.92

3

s.d.

3.39

3.78

3.34

9.262.51

3.35

Alphas

.53

.80

.93

.52

'N = 24.


Importantly, the measurement of archetypes does not (and should not)involve measuring the commitment of individuals within organizations toan archetype—hence our emphasis on the use of interviews to producedescriptive accounts of organizational arrangements. Structures and systemsreflect and embody ideas and values. To the extent that structures and sys-tems consistently embody one set of ideas and values, an observer can talkof archetype coherence. But current organization members need not be com-mitted to those ideas and values (Greenwood & Hinings, 1988). This distinc-tion is important methodologically: The pattern of commitment within anorganization to particular interpretive schemes and their organizational im-plications is a dynamic of change, not a definitional component of an ar-chetype. The existence of an archetype is established not by measuring theideas, beliefs, and values of organization members, but by comparing theempirically revealed pattern of the structures and systems of an organizationto the analytical pattern of a previously identified ideal template.

RESULTS

Hypothesis 1

Table 5 summarizes the organizational arrangements of the 24 organiza-tions at the three different survey points. In the first period, the organizationswere significantly less likely to have archetypal status (only 33 percent didso) than in the two subsequent periods, when 42 and 71 percent, respec-tively, conformed to archetypes. These statistics neither clearly support nordisprove Hypothesis 1.

Importantly, however, cross-sectional data such as appear in Table 5 arenot satisfactory alone because they miss broad patterns of intellectual move-ment occurring across the population of organizations. Results for any oneyear cannot indicate whether the survey point is typical or reflects a transi-tional and atypical state. And even three waves of statistics require inter-pretation. We can illustrate this point by recounting the institutional dy-namics of the period in question.

At the beginning of the research inquiry, both design archetypes—or,more accurately, the ideas and values underpinning them—formed part ofthe institutional sector's currency and debate. The heteronomous profes-sional bureaucracy had the longer history and was the predominant model

TABLE 5Archetype Congruence at Three Survey Points

Survey Points

1974-751977-781981-82Total

Number

8101735

Congruence

Percentage

33427149

Absence of Congruence

Number

16147

37

Percentage

67582951


before 1967. During the early 1970s, a series of reports criticized the failureof local organizations to handle serious and growing problems, such as innercity decline, economic deterioration, and social and educational disadvan-tage. The structures and systems of these organizations, with their profes-sional basis, were singled out as a major obstacle to more flexible and inte-grated responses. In addition, the worsening fiscal context, in which thecentral government coerced local organizations to cut expenditures,strengthened the intellectual and political position of those advocating theadoption of alternative organizational arrangements.

The pressure to adopt corporate structures and processes was thus con-siderable. Indeed, such was the growing legitimacy of the corporate modelthat a report prepared and endorsed by the central government extolled thevalues of corporate operations and produced charts depicting corporatestructures; charts of systems were not offered, although there was discussionin the body of the report of new policy planning and human resource ap-proaches. The important point is that the flux and ferment of ideas wasunusually high in the early years of the research period and contributed tothe relative instability of organizational archetypes in the period pictured inTable 5.

The results reported here are consistent with the complementary ana-lytical thrusts of Tushman and Romanelli (1985) and Zucker (1983, 1987).Tushman and Romanelli suggested that organizations evolve through a se-ries of lengthy stages of design convergence punctuated by dramatic andconcentrated upheavals, or metamorphoses. Tushman and Romanelli's the-sis is in terms of individual organizations rather than populations or classesof organizations, but Zucker's work indicates that new forms of organiza-tional arrangements, once legitimated within an institutional sector, are rap-idly adopted by most organizations within the sector, much in the way thatstrategic recipes (Child & Smith, 1987) spread throughout an industry. Thisscenario seems to be what occurred for the sample of organizations exam-ined here. Prior to 1967, few of the organizations operated within the cor-porate bureaucracy archetype, and few had reviewed their overall designssystematically (Creenwood, Norton, & Stewart, 1969). However, by 1972 thecorporate archetype had been articulated and legitimated as the appropriatemodel—hence the apparent instability reported in Table 5 and the move-ment toward the archetype. The appropriate conclusion to be drawn fromthe data shown in Table 5 is, therefore, that contextual factors were pushingorganizations out of one set of coherent structures toward another. Such aconclusion leads to a further observation, which is that the extent to whichadherence to archetypes is the normal organizational practice has to reflectthe degree of turbulence or stability within the institutional context. Theflux and ferment of ideas in the sector examined here was unusually high inthe early stages of the research and contributed to the low proportion ofcoherently organized organizations in 1974-75. The later data, those gath-ered at the third survey point, support Hypothesis 1.


Hypothesis 2

If the proposition that organizations alter discordant structures toachieve archetypal status is valid, the 16 organizations that lacked designcoherence at the beginning of the survey period should have been seen tomove toward a state of coherence. The relevant data are summarized inFigure 2, which shows the tracks (Greenwood & Hinings, 1988), or historicalpaths, of the 24 organizations across the three survey points. A track is thetemporal association of an organization with one or more archetype. It isabout the extent to which an organization remains over time within theassumptions and parameters of a given archetype or moves between arche-types (Hinings & Greenwood, 1988a). Of the 16 organizations that initiallylacked coherence, 9 (56%) attained it. Further, 11 (61%) organizations thatlacked coherence at the first and second surveys moved in the predicteddirection (tracks 2,3, and 5 of Figure 2). Thus, results reasonably support theidea of organizations evolving from incoherence to coherence.

Hypothesis 3

Figure 2 can also be used to examine the extent of inertia. The degree ofabsolute inertia is low, in that only 6 organizations remained within thesame archetype throughout the research period (track 1). A further 4 organ-izations exhibited relative inertia, beginning either within, or close to anarchetype before moving into it, and staying within it over the subsequentsurvey periods (track 2 and, in the case of two organizations, track 3). Thus,a total of 10 organizations can be classified as exhibiting inertia. Of these,only 2 started as heteronomous professional bureaucracies and stayed there.The other 8 were all in or close to the corporate bureaucracy archetype, sotheir inertia was legitimate.

The degree of inertia shown by these 10 organizations can be contrastedwith the small extent of movement between archetypes. Only 2 organiza-tions reoriented from one archetype to another between 1974 and 1982.Given that we predicted that inertia would occur after archetype coherencewas established, we can conclude that 10 of 12 eligible organizations (83%)exhibited inertia. (Of the remaining 12 organizations, 7 did not attain coher-ence, and 5 attained coherence only at the third survey point.) The 5 thatattained coherence were all within the corporate bureaucracy archetype.

These figures indicate the existence of dynamics that hold in place aprevailing set of assumptions and design arrangements, once those arrange-ments are coherent. The strength of the dynamics and their influence maynot be absolute, but they exist and work toward archetypal stability. Hy-pothesis 3 thus has some validity, although it should be borne in mind thatthe majority of organizations studied here practiced relative rather than ab-solute inertia.

DISCUSSION

The starting point for this study was the idea that organizational designarrangements can usefully be considered as wholes rather than as collections


DO

c

oa.

IM

C..2CB _N C

'3 '5S cu

CM

u

2H

ocu>>CB

(6)

sc0

CD

coh

QJCCC

Q}

u

03

oCJ

"•

CO01

Qi

QJCJCQ3

(Dco

h

-aQ3

cCC

QJ

Di

CJ isj '^ • .

QJ Crt

U 03

O

(Jffl

03CJ

cQJ

coh

X!

ffl

Coh

eren

ce(8

org

aniz

atio

ns)

Lac

k of

coh

eren

cepe

rsis

ts (1

2)

Lac

k of

coh

eren

ce(1

6 or

gani

zati

ons)

co

IcCC00

O

s?oc

•a


of disparate elements and that underlying sets of ideas, beliefs, and valuesgive these wholes pattern. A set of structures and systems congruent with aninterpretive scheme constitutes an archetype.

The results presented are not clear-cut. We are reasonably satisfied thatorganizations do tend to operate with structures and systems that approxi-mate archetypes (Hypothesis 1); however, the evidence is less than defini-tive because, we have suggested, of the flux of circumstances in the focalsector at the time of the research. There is also evidence that organizationstend to move toward archetype coherence (Hypothesis 2). And the dataconfirm Hypothesis 3, predicting that major organizational change, or pas-sage between archetypes, is less common than is inertia—archetype stabil-ity.

These results are not to be thought of as disappointing because they areless than clear-cut. The process of uncovering and understanding themraises important issues and themes. Notably, the research highlights theimportance of researchers' incorporating institutional and temporal contextswithin their analytical frameworks. Furthermore, the importance of thesecontextual factors begets certain methodological considerations.

The Importance of Institutional Context

There is no doubt that organizations exist in a milieu of ideas aboutappropriate ways of succeeding and behaving in their task environments.Furthermore, greater and lesser networks facilitating the dissemination andpromotion of specific recipes will characterize different institutional arenas(Hinings & Greenwood, 1988b; Whipp & Clark, 1986]. At present, there israther more evidence of how institutionally located frameworks of ideas andvalues influence not-for-profit organizations than similar evidence from theprivate sector; Zucker (1988), for instance, contains examples predomi-nantly from the not-for-profit sector. But there are interesting examples inthe private sector of how an organization's institutional environment shapescore values and provides prescriptions for appropriate patterns of organizedactivity (e.g., Baum & Oliver, 1991; Child & Smith, 1987; Greenwood, Hin-ings, & Brown, 1990; Pettigrew, 1985; Worthy, 1985).

In general, however, comparatively little is known of the archetypesused in different institutional arenas. The literature provides suggestionsand clues, but a misleading concern for generic organizational types stilldominates the field. Admitting the importance of underpinning purposivevalues specific to institutional sectors challenges the case for generic types.Our research, because of its setting in a single institutional environment,does not demonstrate that generic types do not exist. But we would recom-mend that greater effort go into recognizing and uncovering which arche-types are to be found in which institutional arenas, and why.

The Importance of Temporal Context

The study of organizational archetypes has to be firmly set in a temporalcontext. This does not simply mean the adoption of a longitudinal frame of


reference, although that is a clear methodological implication. Encompass-ing the temporal context is also a matter of appreciating the trends occurringat the time research is conducted. Thus, the present work indicated thestrong prescription in favor of the corporate hureaucracy archetype, and wehave heen careful to interpret the degrees of inertia and change demon-strated in light of that trend. Researchers have to interpret organizations withknowledge of the degree to which the temporal context supports conver-gence or, alternatively, prescrihed metamorphosis and of the strength anddirectional content of pressures for change affecting the population within agiven institutional arena. Neither individual organizations nor populationsof organizations can he studied as though they have no past or direction.

We further suggest that temporal context should have a hiographicalcomponent (Kimherly, 1987]. Understanding the degree and importance ofchange in an organization's arrangements and the extent to which thosechanges do or do not indicate a loosening or tightening of attachment toparticular archetypes is difficult unless a researcher appreciates not only thebroad sweep of prescriptive influences hut also the historical position of theindividual organization. For us, this meant understanding whether an organ-ization was historically a professional or corporate hureaucracy. Elsewhere,it has meant understanding whether an organization has a history of inertiaor of change, whether senior managers have been recruited from within oroutside the organization, and whether performance has traditionally heenperceived as deficient or strong (Hinings & Greenwood, 1988b]. The point isthat change cannot be examined without some attention to the biography ofthe organization under study.

Raising the importance of the biographical context is very close to ex-plicitly raising the importance of understanding the dynamics of strategicchange—the forces that shape and facilitate or hinder the degree and direc-tion of an organization's temporal attachment to a particular archetype. Theuncovering of archetypes is thus important hut leads to a set of intriguingquestions. What are the mechanisms by which archetypal recipes are devel-oped and learned within an institutional setting? To what extent and howare archetypes disseminated across institutional boundaries? What factorsencourage inertia and what are the dynamics of archetypal reorientation?

Various researchers are posing and addressing these questions. In earlierwork (Greenwood & Hinings, 1988] we elaborated a language for classifyingsuch movements between archetypes. Tushman and Romanelli (1985] andMiller (1991] have theorized about the pace and phasing of movements. Invery different ways, Hannan and Freeman (1977], GarroU (1984], Tushmanand Romanelli (1985], Pettigrew (1985], and another earlier work (Hinings &Greenwood, 1988a] have provided ideas about the dynamics of archetypechange. Zucker (1987], Ghild and Smith (1987], Galori, Johnson, and Sarnin(1991], and Ghild (1988] have examined the role of institutional context inthe dissemination of ideas and approaches. In other words, researchers haveshown considerable interest in exploring the dynamics and processes oforganizational attachment to, and decoupling from, available design arche-


types. As yet, however, no definitive set of answers has emerged, and inquirypromises to be competitively stimulating.

We have suggested (Hinings & Creenwood, 1988a) that the dynamics ofstability and change can best be understood through a framework dealingwith the interaction of situational context, interpretive schemes, interests,dependencies of power, and organizational capability. Stability will resultwhen there is a fit between these elements and a particular organizationalarchetype, and change will result when contradictions, incompatibilities,and tensions occur between them. We also distinguished between "precip-itating" and "enabling" dynamics. For example, value commitments andinterests, which are situational factors, can precipitate change or enhancestability; power and capability enable change or, again, reinforce stability.

Defining an archetype in terms of structures and systems underpinnedby values and ideas (interpretive schemes), as we have done, directs atten-tion to a key potential dynamic of change, namely, the extent to whichcurrent organizational actors are committed to prevailing and alternativeinterpretive schemes. Four possible patterns of commitment can be envis-aged: (1) widespread commitment to an existing interpretive scheme (a sta-tus quo commitment), (2) widespread commitment to an alternative inter-pretive scheme (a reformative commitment), (3) substantial commitment totwo or more interpretive schemes (a competitive commitment), and (4) lowcommitment to prevailing and alternative interpretive schemes (an indiffer-ent commitm^ent).

In the case of a status quo commitment, pressure toward organizationalinertia will be strong. Reformative commitments are liable to lead to inter-archetype movement. Competitive commitments destabilize organizationalarrangements without necessarily activating complete reorientations, pro-ducing archetype incoherence. Indifferent com^mitments can be associatedwith inertia but are amenable to externally induced change.

The central point is that understanding the key link between patterns ofstructures and systems and the underlying interpretive scheme provides aresearcher with three things essential for theorizing about change. First,archetypes provide a definitional basis that allows classification of the scaleof change; for instance, change can be defined as incremental or quantum.Second, archetypes offer an understanding of the consistency or inconsis-tency of the direction of change as it occurs. Third, a critical dynamic ofchange and inertia is identified, namely, the pattern of commitment to theinterpretive scheme embodied in an organization's structures and systems.

Methodological Implications

The language of analysis in this article—inertia, momentum, evolution,reorientation, tracks—is about motion and change. The study of archetypesis thus the study of organizations over time. Interestingly, although the re-search period reported upon here was considerably longer than that usuallyfound in organization theory research, we have concluded that the periodwas too short because it was difficult to control the effects of broad institu-


tional developments. A fourth wave of data collection would have allowedus to arrive at stronger conclusions about all three hypotheses.

A second methodological lesson to be derived from the present researchis that an understanding of organizational structures and systems and oforganizational change requires a serious effort to systematically uncover theideas, beliefs, and values of organization members and the ideas and valuescontained within the focal institutional and temporal contexts. Organiza-tional patterns are not disembodied attributes of organizations but reflectorsof meanings. Therefore, archetypes can be uncovered only through a processof research in which researchers seek to assimilate the intellectual milieus ofthe temporal and institutional contexts of the organizations they wish tounderstand. The signal implication of defining archetypes in terms of valuesand ideas and of acknowledging the possibility of those values and ideashaving institutional specificity is that the process of intellectual assimilation(understanding the form and implication of ideas and values) is possibleonly through a variant of the grounded approach deployed here. The arche-types discussed were not theoretical constructs applied to the institutionalpopulation but constructs derived from the population itself. It was, more-over, striking to us that the interpretation of the more quantitatively basedanalysis of the second stage of the research was possible only by virtue of thelengthy (three-year) process of a grounded approach tolerated in stage one.

One difficulty that could be encountered in research of the kind de-scribed here and be exacerbated as a research time frame lengthens is theemergence during the research effort of unanticipated archetypes. Thelonger the period of research, the greater the chance that another phase ofintellectual ferment and upheaval will occur. Thus, in the institutional set-ting of the present study, toward the end of the project ideas about privati-zation and total quality were being discussed and subsequently would affectdefinitions of domain, modes of organization, and appropriate criteria ofperformance in the institutional sector of interest here. The study of organ-izational change as the movement within and between archetypes, whichpromises to be theoretically rewarding, may well also be methodologicallychallenging and frustrating, and it may require some creativity.

Conclusions

The purpose of the present study was threefold: first, to examine theconcept of archetype in order to provide a satisfactory definition; second, toassess the concept's empirical relevance; and third, to review some of thetheoretical and methodological issues involved in the study of archetypes.The definition provided stresses the need to adopt a holistic approach to thestudy of organizational arrangements and to recognize interpretive schemesas providing the basis of pattern. The findings give reasonable support to theideas that archetypes exist as intellectual constructs within institutionalsectors and that organizations tend to seek to organize their structures andsystems in terms of those archetypes. Certainly, there is sufficient encour-agement from the data to warrant further exploration and utilization of ar-


chetypes as a focus of theoretical and research effort. These modest pur-poses, we stressed at the outset, are within the traditions of the long andimportant search within organization theory for typologies and taxonomies,a search integral to, if sometimes implicit in, much contemporary work.

In addressing the third purpose—to uncover relevant theoretical andmethodological issues—we have raised important issues for subsequent re-search efforts aimed toward understanding organizational change. In partic-ular, we have argued for inclusion of the institutional and temporal contextsand for the adoption of methodologies more sensitive to the institutionalspecificity of meanings and the purpose at the core of the archetype concept.Our methodological recommendations imply a style of research demandingin terms of resources and the commitment of time, but we believe that thetheoretical potential outweighs the methodological price.

REFERENCES

Abernathy, W. J., & Clark, K. B. 1985. Innovation: Mapping the winds of creative destruction.Research Policy, 14: 3-22.

Andrews, K. R. 1980. The concept of corporate strategy. Homewood, IL: Irwin.

Astley, W. G. 1985. The two ecologies: Population and community perspectives on organiza-tional evolution. Administrative Science Quarterly, 30: 224-241.

Barley, S. R. 1986. Technology as an occasion for structuring: Evidence from observations of CTscanners and the social order of radiology departments. Administrative Science Quar-terly, 31: 78-108.

Bartunek, J. M. 1984. Changing interpretive schemes and organizational restructuring: The ex-ample of a religious order. Administrative Science Quarterly, 29: 355-372.

Baum, J., & Oliver, C. 1991. Institutional linkages and organizational mortality. AdministrativeScience Quarterly, 36: 187-218.

Bennis, W. G., Benne, K. D., & Chin, R. 1985. The planning of change. New York: Holt, Rine-hart & Winston.

Betton, J., & Dass, G. 1985. The application of population ecology models to the study oforganizations. Academy of Management Review, 10: 750-757.

Beyer, J. 1981. Ideologies, values, and decision making in organizations. In P. C. Nystrom &W. H. Starhuck (Eds.), Handbook of organizational design, vol. 2: Remodeling organiza-tions and their environments: 166-202. New York: Oxford University Press.

Blau, J., & McKinley, W. 1979. Ideas, complexity and innovation. Administrative ScienceQuarterly, 24: 200-219.

Blau, P. M., & Schoenherr, R. A. 1971. The structure of organizations. New York: Basic Books.

Callon, M., & VignoUe, J. P. 1976. Organisation locale et enjeux societaus. Sociologie du travail,3(2): 233-255.

Calori, R., Johnson, G., & Sarnin, P. 1991. French and British top managers' understanding of thestructure and the dynamics of their industries: A cognitive analysis and comparison. Rrit-ish Joumai of Management, 3: 61-78.

Carroll, G. R. 1984. Organizational ecology. In R. H. Turner & J. F. Short (Eds.), Annual reviewof sociology, vol. 10: 71-93. Palo Alto, CA: Annual Reviews.

Champion, D. J. 1975. The sociology of organizations. New York: McGraw-Hill.

1078 Academy 0/ Management Journal October

Child, J. 1973. Predicting and understanding organization structure. Administrative ScienceQuarterly, 18: 168-185.

Child, J. 1988. On organizations in their sectors. Organization Studies, 9: 13-19.

Child, J., & Smith, C. 1987. The context and processes of organizational transformation. Journalof Management Studies, 24: 565-593.

Clegg, S. 1975. Power, rule and domination. London: Routledge & Kegan Paul.

Cummings, L. L. 1984. Compensation, culture, and motivation: A systems perspective. Organ-izational Dynamics, 12(1): 33-44.

Daft, R., & Macintosh, N. 1984. The nature and use of formal control systems for managementcontrol and strategy implementation. Journal of Management, 10(1): 43-66.

Dearlove, J. 1979. The reorganization of Rritish local government Cambridge: CambridgeUniversity Press.

DiMaggio, P., & Powell, W. 1983. The iron cage revisited: Institutional isomorphism and col-lective rationality in organizational fields. American Sociological Review, 48: 147-160.

Doz, Y., & Prahalad, V. 1988. A process model of strategic redirection in large, complex firms.In A. Pettigrew (Ed.), The management of strategic change: 63-83. Oxford: Basil Black-well.

Dyer, L. 1983. Bringing human resources into the strategy formulation process. Human Re-source Management, 22: 257-272.

Galhraith, J. 1975. Organizational design. Reading, MA: Addison-Wesley.

Creenwood, R. 1984. Incremental budgeting: Antecedents of change. Joumal of Public Policy,4: 277-306.

Greenwood, R., & Hinings, C. R. 1988. Design archetypes, tracks and the dynamics of strategicchange. Organization Studies, 9: 293-316.

Creenwood, R., Hinings, C. R., & Brown, J. 1990. The P2-form of strategic management: Corpo-rate practices in the professional partnership. Academy of Management Joumal, 33: 725-755.

Creenwood, R., Norton, A. L., & Stewart, J. D. 1969. Recent changes in the internal organizationof county boroughs: Part 1: committees. Public Administration, Summer: 151-167.

Creenwood, R., & Stewart, J. D. 1974. Corporate planning in English local government: Ananalysis with readings 1967-72. London: Charles Knight.

Hannan, M., & Freeman, J. 1977. The population ecology of organizations. American Journal ofSociology, 85: 929-966.

Hinings, C. R., & Creenwood, R. 1988a. The dynamics of strategic change. Oxford: Basil Black-well.

Hinings, C. R., & Creenwood, R. 1988b. The normative prescription of organizations. In L.Zucker (Ed.), Institutional patterns and organizations: 53-70. Cambridge, MA: Ballinger.

Huff, A. S. 1982. Industrial influences on strategy reformation. Strategic Management Joumal,3: 119-131.

Johnson, C. 1987. Strategic change and the management process. Oxford: Basil Blackwell.

Karpik, L. 1978. Organizations, institutions and history. In L. Karpik (Ed.), Organization andenvironment: Theory, issues and reality: 15-68. Beverly Hills, CA: Sage.

Khandwalla, P. N. 1977. The design of organizations. New York: Harcourt Brace Jovanovich.

Kikulis, L., Slack, T., & Hinings, C. R. 1992. Institutionally specific design archetypes: A frame-work for understanding change in national sport organizations. International Review forthe Sociology of Sport, 27: 343-370.


Kilmann, R., Saxton, M., & Serpa, R. (Eds.). 1985. Gaining control of the corporate culture. SanFrancisco: Jossey-Bass.

Kimberly, J. R. 1987. The study of organization: Toward a biographical perspective. In J. W.Lorsch [Ed.], Handbook of organizational behavior: 223-237. Englewood Cliffs, NJ: Pren-tice-Hall.

Kimberly, J. R., & Quinn, R. E. (Eds.). 1984. Managing organizational transitions. Homewood,IL: lrwin.

Lammers, C. 1978. The comparative sociology of organizations. In R. H. Turner, J. Coleman, &R. C. Fox (Eds.), Annual review of sociology, vol. 4: 485-510. Palo Alto, CA: AnnualReviews.

Lawrence, P., & Lorsch, J. G. 1967. Organization and environment. Cambridge, MA: HarvardUniversity Press.

Lindblom, C. 1958. The science of "muddling through." Public Administration Review, 29:79-88.

McKelvey, W. 1975. Guidelines for the empirical classification of organizations. Administra-tive Science Quarterly, 20: 509-525.

McKelvey, W. 1982. Organizational systematics: Taxonomy, evolution, classification. Berke-ley: University of California Press.

McKinney, J. C. 1966. Constructive typology and social theory. New York: Appleton-Century-Crofts.

Meyer, J., & Rowan, B. 1977. Institutional organizations: Formal structure as myth and cere-mony. American Journal of Sociology, 83: 440-463.

Miles, R., & Snow, G. 1978. Organizational strategy, structure and process. New York: Mc-Graw-Hill.

Miller, D. 1981. Towards a new contingency approach: The search for organizational gestalts.Journal of Management Studies, 18: 1-26.

Miller, D. 1987. Strategy making and structure: Analysis and implications for performance.Academy of Management Journal, 30: 7-32.

Miller, D. 1991. The Icarus paradox. New York: Harper Gollins.

Miller, D., & Friesen, P. 1980. Momentum and revolution in organization adaptation. Academyof Management Journal, 23: 591-614.

Miller, D., & Friesen, P. 1984. Organizations: A quantum view. Englewood Cliffs, NJ: Prentice-Hall.

Mintzberg, H. 1979. The structure of organizations. Englewood Cliffs, NJ: Prentice-Hall.

Mintzberg, H. 1983. Structure in fives: Designing effective organizations. Englewood Cliffs,NJ: Prentice-Hall.

Nadler, D. A. 1987. The effective management of organizational change. In J. W. Lorsch (Ed.),Handbook of organizational behavior: 358-369. Englewood Cliffs, NJ: Prentice-Hall.

Nadler, D. A., & Tushman, M. 1989. Organizational frame bending: Principles for managingTeoTientation. Academy of Management Executive, 1: 194-204.

Oliver, C. 1991. Strategic responses to institutional processes. Academy of Management Re-view, 16: 145-179.

Ouchi, W. 1977. The relationship between organizational structure and organizational control.Administrative Science Quarterly, 22: 95-113.

Perrow, C. 1982. Complex organizations: A critical analysis. Glenview, IL: Scott, Foresman.


Pettigrew, A. 1973. The politics of organizational decision making. London: Tavistock.

Pettigrew, A. 1985. The awakening giant. Oxford: Basil Blackwell.

Pfeffer, J. 1981. Power in organizations. Boston: Pitman.

Pinder, C. C, & Moore, L. R. 1979. The resurrection of taxonomy to aid the development ofmiddle-range theories of organizational behavior. Administrative Science Quarterly, 24:99-118.

Porter, M. E. 1980. Competitive strategy. New York: Free Press.

Pugh, D. S., Hickson, D. J., & Hinings, C. R. 1969. An empirical taxonomy of work organizations.Administrative Science Quarterly, 14: 115-126.

Quinn, J. B. 1982. Strategies for change: Logical incrementalism. New York: Irwin.

Ranson, S., Hinings, C. R., & Greenwood, R. 1980. The structuring of organizational structures.Administrative Science Quarterly, 25: 1-17.

Sathe, V. 1985. Culture and related corporate realities. Homewood, IL: Irwin.

Schein, E. 1986. Organizational culture and leadership. San Francisco: Jossey-Bass.

Schuler, R. S. 1984. Personnel and human resource management (2d ed.). St. Paul: West.

Scott, W. R. 1965. Reactions to supervision in a heteronomous professional organization. Ad-ministrative Science Quarterly, 10: 65-81.

Scott, W. R. 1987. The adolescence of institutional theory. Administrative Science Quarterly,32: 493-511.

Singh, J. V., House, R. J., & Tucker, D. J. 1986. Organizational change and organizational mor-tality. Administrative Science Quarterly, 31: 587-611.

Skelcher, C. 1980. From programme budgeting to policy analysis. Public Administration, 59:155-172.

Spender, J. C. 1980. Strategy making in business. Unpublished doctoral dissertation. School ofBusiness, University of Manchester.

Stewart, J. D. 1971. Management in local government: A viewpoint. London: Charles Knight.

Stonich, P. J. 1984. The performance measurement and reward system: Critical to strategicmanagement. Organization Dynamics, 12(1): 45-57.

Thompson, K. 1975. Religious organizations. In J. B. McKinlay (Ed.), People processing: 1-40.New York: Holt, Rinehart.

Tolbert, P. S., & Zucker, L. G. 1983. Institutional sources of change in the formal structure oforganization: The diffusion of civil service reform, 1880-1935. Administrative ScienceQuarterly, 28: 22-39.

Tushman, M. L., & Anderson, P. 1988. Technological discontinuities and organization environ-ments. In A. Pettigrew (Ed.), The management of strategic change: chap. 3. Oxford: BasilBlackwell.

Tushman, M. L., & Romanelli, E. 1985. Organizational evolution: A metamorphosis model ofconvergence and reorientation. In L. L. Cummings & B. M. Staw (Eds.), Research in organ-izational behavior, vol. 7: 171-222. Greenwich, CT: JAI Press.

Van de Ven, A., & Drazin, R. 1985. The concept of fit in contingency theory. In L. L. Cummings& B. M. Staw (Eds.), Research in organizational behavior, vol. 7: 333-365. Greenwich,CT: JAI Press.

Walsh, K., Hinings, C. R., Greenwood, R., & Ranson, S. 1981. Power and advantage in organ-izations. Organization Studies, 2: 131-152.

Weber, M. 1947. The theory of social and economic organization. New York: Free Press.


Whipp, R., & Clark, P. 1986. Innovation and the auto industry. London: Frances Pinter.

Wildavsky, A. 1984. The politics of the budgetary process. Boston: Little, Brown.

Worthy, ]. 1985. Shaping an American institution: Bobert E. Wood and Sears Boebuck.Urbana: University of Illinois Press.

Yin, R. K. 1984. Case study research: Design and methods. Beverly Hills, CA: Sage.

Zucker, L. 1983. Organizations as institutions. In S. B. Bacharach (Ed.), Besearch in the soci-ology of organizations, vol. 2: 1-47. Greenwich: JAI Press.

Zucker, L. 1987. Normal change or risky business: Institutional effects of the "hazard" of changein hospital organizations, 1959-1979. Journal of Management Studies, 24: 671-700.

Zucker, L. (Ed.). 1988. Institutional patterns and organizations. Chicago: Ballinger.

Royston Greenwood is the AGT Professor in the Department of Organizational Analy-sis, Faculty of Business, at the University of Alberta. He received his Ph.D. degree fromthe University of Birmingham, England. His current research concerns the managementof professional service organizations, strategic organizational change, and the institu-tional specificity of design archetypes.

C. R. Hinings is the Thornton A. Graham Professor of Business in the Department ofOrganizational Analysis, University of Alberta. His research interests are the manage-ment and dynamics of strategic organizational change, with particular reference toprofessional service firms.

Documents

UNDERSTANDING STRATEGIC CHANGE: THE CONTRIBUTION …bigideasproject.wikispaces.com/file/view/greenwood1993.pdf · ® Academy of Managemenl JournoJ 1993, Vol. 36, No. 5, 1052-1081