Structure design and

470 views

Published on

management

Published in: Design
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
470
On SlideShare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
5
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Structure design and

  1. 1. The Strategic Decision Process and Organizational StructureAuthor(s): James W. FredricksonReviewed work(s):Source: The Academy of Management Review, Vol. 11, No. 2 (Apr., 1986), pp. 280-297Published by: Academy of ManagementStable URL: http://www.jstor.org/stable/258460 .Accessed: 30/12/2012 15:04Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.. Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy of Management Review.http://www.jstor.org This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  2. 2. ? Academy of Management Review, 1986, Vol. 11, No. 2, 280-297. The Strategic Decision Process and Organizational Structure JAMESW. FREDRICKSON Columbia University Contributions from the strategic decision process literature are synthe- sized and integrated with literature on organizational structure. Prop- ositions emerge that describe how the characteristics of an organiza- tions strategic decision process are affected by its structure. Also discussed are the patterns of strategic process characteristics that are likely to be associated with different types of structures. Conclusions are reached on issues such as the accuracy of alternative models of the strategic decision process, and the appropriate unit of analysis for studying that process. For many years authors have suggested that ture. It also attempts to encourage a broader de-the relationship between organizational strategy bate, by assuming a perspective that is in sharpand structure is reciprocal. Only recently has contrast to both the "structure follows strategy"there been widespread agreement that structure view and work regarding the effect that environ-can have a profound impact on strategy through ment and other variables may have on struc-its direct effect on the strategic decision-making ture. The paper begins with a review of litera-process (Bourgeois & Astley, 1979; Burgelman, ture that traces the strategy/structure debate,1983; Fahey, 1981). A variety of strategic pro- identifies important characteristics of the stra-cess and structural variables have been used tegic decision process, and describes those di-in describing isolated aspects of this relationship, mensions of structure that are most likely to affectand competing explanations have been pro- strategic decision making. The second sectionvided. However, most of this work remains frag- draws on contributions from several areas tomented and major theoretical gaps persist. develop propositions that describe how the char- This paper addresses the above problems by acteristics of an organizations strategic decisionsynthesizing and integrating previous work; it process are affected by individual dimensions ofalso offers new explanations to fill critical gaps. structure. In the final section, a link is establishedThis is a "first step" to encourage investigation between each of the previously discussed dimen-and debate on how an organizations strategic sions and Mintzbergs (1979) well-known struc-decision process is affected by its formal struc- tural "types." This makes it possible to describe the pattern of strategic process characteristics Donald C. Hambrick, R. T. Lenz, William H. Newman, that is likely to occur in each type, and to under-Lloyd E. Sandelands, and Michael L. Tushman provided stand why different structures are typically morevaluable comments on an earlier version of this manuscript. successful in different contexts. Requests for reprints should be sent to James W. Fred- The arguments presented here lead to severalrickson, Graduate School of Business, Columbia University, conclusions. Among them are that the accuracyNew York, NY 10027. of alternative models of the strategic decision- 280 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  3. 3. making process (Allison, 1971; Mintzberg, 1973), zationally rational outcomes" in spite of theiras well as the appropriate "unit of analysis" for members cognitive limitations (Simon, 1976). Itunderstanding that process, vary with the type also helps management to control the decision-of structure. making environment and facilitate the process- ing of information. This link is apparent in Strategic Process and Structure Bowers (1970)comment that "when management chooses a particular organization form, it is pro-Historical Relationship viding not only a framework for current opera- In studying the development of Americas tions but also the channels along which strate-dominant industrial organizations, Chandler gic information will flow . . ." (p. 287).(1962) observed that major increases in unit Organizations often have some units whosevolume, geographic dispersion, and vertical and structures are different from that which charac-horizontal integration were eventually followed terizes the organization as a whole (e.g., whileby changes in structural form. Although a new the organization is generally decentralized, somestructure was adopted only after a change in units may be very centralized). However, thisstrategy made its predecessor dysfunctional, the paper is concerned with the structure that bestlogic of a relationship was compelling. In ad- describes the whole organization, a concept thatdition, several early studies confirmed an asso- will be referred to as its "dominant" structure.ciation between these variables (Fouraker & Therefore, it is argued that the characteristics ofStopford, 1968; Rumelt, 1974), and the proposi- a firms strategic decision-making process aretion that "structure follows strategy" became affected by its overall, dominant structure. Thiswidely accepted. will undoubtedly seem like an ill-founded asser- In spite of the widespread acceptance of the tion if one believes that (a) all strategic decisionsabove relationship, there is a growing body of are made by one or a very few top-level execu-literature that suggests that there is a major effect tives, or that (b) such decisions are made outsidefrom structure to strategy (i.e., once a structure of the dominant structure. As a result, it is impor-is in place it will influence a firms strategic deci- tant to illustrate that in many instances it is thesesion process, and ultimately its strategy). For beliefs that are ill-founded.example, Bower (1970) characterized structure Regarding the first issue, it should be recog-broadly, as the context within which decisions nized that "choice" is only one of many activitiesare made, and observed that ". . structure may that are involved in the decision-making processmotivate or impede strategic activity. . ." (p. 67). (e.g., information search). Moreover, numerousNumerous other contributors (Bobbitt & Ford, authors (Crozier, 1964; Mintzberg, 1979; Simon,1980; Duncan, 1979; Hedberg, Nystrom, & Star- 1976;Thompson, 1967)agree that only in the mostbuck, 1976; Jelinek, 1977) have argued simply simple of organizations are all of the activitiesthat structure constrains strategic choice. controlled by one individual. Therefore, because To understand why it is logical for the strate- it is difficult to obtain and comprehend all of thegic decision process to be affected by structure, information that is needed to make strategic deci-one must understand the relationship between sions in a large organization (Quinn, 1980), thedecision making and structure. March and Simon strategic process typically requires contributions(1958) get to the heart of this relationship by from people with a wide range of expertise andarguing that an organizations structure imposes from numerous levels (Carter, 1971; Crozier,"boundaries of rationality" that accommodate 1964).members cognitive limitations. By delimiting The second issue that warrants clarificationresponsibilities and communication channels, concerns the strategic impact of a firms domi-structure allows organizations to achieve "organi- nant structure. It is recognized that organizations 281 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  4. 4. may deal with strategic decisions by creating The stream of work identified above concludedtask forces, committees, and project teams that the dominant models of strategy formula-(Thompson, 1967) and by using nominal group tion differ on the following six characteristics: (a)(Delbecq, Van de Ven, & Gustafson, 1975) or process initiation, (b) the role of goals, (c) therelated techniques. However, the very nature of means/ends relationship, (d) the explanation ofthe strategic process makes it unlikely that such strategic action, (e) the comprehensiveness ofmechanisms can fully negate the impact of the decision making, and (f) comprehensiveness indominant structure. For example, Mintzberg integrating decisions. A description of these char-(1979) has argued that a strategic issue can acteristics is provided in Table 1, which also"emerge" from anywhere in an organization; it identifies critical questions regarding each. Al-is not necessarily recognized first by those at the though these six characteristics and their accom-top of the organization. Therefore, the issue may panying questions are certainly not an exhaus-become evident to top-level managers, and may tive list, it is suggested that their basis in thebecome the focus of a specially formed task force theoretical literature makes them particularlyor committee, only after it has been filtered important. Therefore, this paper will focus onthrough the organizations dominant structure. how they are affected by organizational struc-Moreover, it is suggested that the likelihood of a ture, following a brief discussion of that topic.firms using such mechanisms can be predictedby the characteristics of its dominant structure(e.g., a firm that is highly centralized is unlikely Structural Dimensions and Typesto create special committees). As the first step in Structure refers to an organizations internalunderstanding precisely how a firms strategicdecision process is affected by its structure, the pattern of relationships, authority, and commu-next section identifies several important charac- nication (Thompson, 1967). It has been character-teristics of that process. ized on a variety of dimensions and illustrated using a variety of "types" (e.g., functional orStrategic Process Characteristics divisional). Moreover, debate continues regard- Most studies of the strategic decision process ing the validity of measures that have been usedhave produced either a very "focused" set of to assess structures dimensions (Blackburn, 1982;observations regarding one process question, or Fry, 1982; Walton, 1981), and the link betweena very rich but "loose" description of the entire the dimensions and types is often ignored. How-decision process. An example of the latter is the ever, three dimensions of structure-centrali-"phases" and "routines" identified by Mintzberg, zation, formalization, and complexity-haveRaisinghani, and Theoret (1976). However, a re- received more attention than any others (Child,cent comparison (Fredrickson, 1983) of the two 1974; Ford & Slocum, 1977; Fry, 1982; Hage &types of models-the synoptic and incremental- Aiken, 1967; Hall, 1977; Van de Ven, 1976) andthat appear most frequently in the strategy for- they appear to have the greatest implications formulation literature, identified several concept- strategic decision making. Each of these dimen-ually distinct though related characteristics on sions is also the dominant characteristic of a well-which they differ. That comparison drew on the known structural type.work of numerous authors, but was based most Centralization refers to the degree to whichdirectly on contributions by Lindblom (1959) and the right to make decisions and evaluate activi-Mintzberg (1973). In examining alternative de- ties is concentrated (Fry & Slocum, 1984; Hall,scriptions of the strategic decision-making pro- 1977). A high level of centralization is the mostcess, those authors reached very similar conclu- obvious way to coordinate organization decisionsions regarding the critical characteristics on making, but it places significant cognitive de-which such processes could be differentiated. mands on those managers who retain authority. 282 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  5. 5. Table 1Critical Strategic Decision Process Characteristics Process Characteristics Description and Questions 1. Process initiation Concerned with how and where the process is initiated. Is the process initiated as a reaction to problems/crises, or the proactive pursuit of opportunities and interests? At what level (operating or strategic) would a stimulus have to appear before members would recognize it as being strategic? What level assumes primary responsibility for initiating the process? 2. Role of goals Issues regarding the role that goals play in the decision process. Will decisions be made to achieve individual versus organization-level goals? Will goals be "remedial" changes from the status quo, or "positive," future intended states? Are the goals likely to be conceptualized in precise versus general terms? 3. Means/ends Concerned with the relationship that exists between means (alternatives) and ends relationship (goals). What is the likelihood that means will displace ends (goals) in the decision process? Will goals persist in the face of significant changes in the available means? 4. Explanation of Considers alternative explanations of the process that resulted in strategic action. strategic action Is strategic action most accurately characterized as intendedly rational strategic choice, the result of standardized organizational processes, an internal process of political bargaining, or some other explanation? What is the likelihood that stragetic moves will be incremental versus major departures from the existing strategy? 5. Comprehensiveness in Attempting to identify the factors that limit the comprehensiveness of the decision making strategic decision process. Is the primary constraint on the comprehensiveness of the strategic process top managements cognitive limitations, the detail achieved in the design of standardized organizational processes, or managers parochial perceptions? 6. Comprehensiveness in Concerned with how comprehensively individual decisions are integrated. What integrating decisions level of integration is achieved to form an overall strategy?Mintzberg (1979) has discussed this issue by sug- cant consequences for organizational membersgesting that an individual does not have the cog- because it specifies how, where, and by whomnitive capacity or information that is needed to tasks are to be performed. A high level of formal-understand all the decisions that face a complex ization has the benefit of eliminating role ambi-organization. Therefore, it is not surprising that guity, but it also limits members decision-makinga negative relationship has been reported be- discretion. Therefore, it is generally argued thattween an organizations size and its degree of the level of formalization must be matched withcentralization (Pugh, Hickson, Hinings, & Turner, the level of professionalism because formaliza-1968). tion threatens professional autonomy (Perrow, The degree of formalization specifies the extent 1972).to which an organization uses rules and proce- Complexity refers to the condition of beingdures to prescribe behavior (Hage & Aiken, 1969; composed of many, usually interrelated, parts.Hall, 1977). Therefore, formalization has signifi- Regarding organizational structure, Hall (1977) 283 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  6. 6. suggests that there are three potential sources of The above arguments suggest that the threecomplexity-horizontal and vertical differentia- dimensions of organizational structure havetion, and spatial dispersion. Therefore, an orga- major implications for decision making. There-nization that simultaneously has numerous fore, in the following discussion each of theselevels, broad spans of control, and multiple geo- conceptually independent dimensions-centrali-graphic locations would be considered highly zation, formalization, and complexity-is dis-complex. While such a structure is often consid- cussed in terms of its likely impact on the strate-ered appropriate for firms that compete in highly gic process characteristics and questions iden-differentiated environments, it is important to rec- tified earlier in Table 1 (i.e., how and where theognize that a high level of complexity makes it process is initiated, the role of goals, and so on).difficult to coordinate and control decision activ- Table 2 summarizes the dimension-specific prop-ities (Lawrence & Lorsch, 1967). ositions that emerge from that discussion.Table 2Propositions Regarding the Effects of Three Dimensions of Structure Centralization Formalization Complexity Propositions 1.A-D. As the level Propositions 2.A-D. As the level Propositions 3.A-D. As the level of centralization increases, so does of formalization increases, so does of complexity increases, so does the probability that- the probability that- the probability that- 1-A. the strategic decision 2-A. the strategic decision 3-A. members initially exposed to process will be initiated only by process will be initiated only the decision stimulus will not the dominant few, and that it will in response to problems or crises recognize it as being strategic, or be the result of proactive, that appear in variables that are will ignore it because of parochial opportunity-seeking behavior; monitored by the formal system; preferences; 1-B. the decision process will be 2-B. decisions will be made to 3-B. a decision must satisfy a oriented toward achieving achieve precise, yet remedial large constraint set, which "positive" goals (i.e. intended goals, and that means will dis- decreases the likelihood that future domains) that will persist place ends (goals); decisions will be made to achieve in spite of significant changes organization-level goals; in means; 1-C. strategic action will be the 2-C. strategic action will be the 3-C. strategic action will be the result of intendedly rational, result of standardized organiza- result of an internal process of "strategic choice," and that moves tional processes, and that moves political bargaining, and that moves will be major departures from the will be incremental; and will be incremental; and existing strategy; and 1-D. top managements cognitive 2-D. the level of detail that is 3-D. biases induced by members limitations will be the primary achieved in the standardized parochial perceptions will be the constraintonthecomprehensiveness organizational processes will be primary constraint on the compre- of the strategic process. The the primary constraint on the hensiveness of the strategic deci- integration of decisions will be comprehensiveness of the strategic sion process. In general, the relatively high. decision process. The integration integration of decisions will be of decisions will be intermediate. low. 284 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  7. 7. Effects of Structures Dimensions be initiated when the innovation interests of a critical mass of coalition members "coalesce" toImpact of Centralization yield proactive behavior. It should be empha- As mentioned earlier, a centralized structure sized that coalition members may not be pro-is one in which the right to make decisions and active (e.g., some managers are very conserva-evaluate activities is concentrated. However, tive) and that the environment may not be recep-because one person seldom controls all strategic tive to such behavior. However, if they do indeedprocess activities even in a highly centralized have proactive interests, their interests are mostorganization (Mintzberg, 1979), in this section it easily pursued when those same members dom-is assumed that decision making rests with a inate the decision process. Therefore, it is sug-small coalition of top-level executives. Such a gested that centralization increases the likeli-view is probably the one that most people have hood that strategic decision making will be a pro-of the strategic decision process, but few have active, opportunity-seeking process.considered how structural centralization affects As implied above, structural centralization alsothe characteristics of that process. The first col- can affect the role that goals play in strategicumn of Table 2 indicates that the strategic impact decision making. For example, under conditionsof centralization is significant, and it first becomes of centralization there are finite limits to theapparent in its effect on how and where the pro- amount of diversity that can exist in the goals ofcess is initiated. coalition members. As a result, it is realistic to As is the case in any organization, members think of members goals as "intended futurethroughout a centralized firm are intermittently domains" (Thompson, 1967) that decisions areexposed to stimuli (e.g., problems and oppor- attempting to achieve. This is in sharp contrasttunities) that have strategic implications for their to an organization that is not centralized, wherefirm. However, with such a structure strategic the diverse preferences of individual membersproblems and opportunities are likely to go un- serve not as goals to be achieved, but as "con-recognized and ignored until they appear be- straints" on the decision process (Simon, 1964). Itfore a coalition member. This is because know- is only when the dominant group is small, as it isledge regarding the likely implications of indi- under conditions of centralization, that strategicvidual stimuli, as well as decision-making decisions are likely to be made with their "posi-authority, is concentrated in the hands of very tive" (i.e., intended future domains) preferencesfew people. Therefore, although centralization in mind.is a means of ensuring that decisions are tightly Simon (1976)has argued that all decisions havecoordinated, it may delay the start of the process fact (means) and value (ends) elements, and thator it may result in a complete failure to respond in a rational model of decision making, ends areto some strategic stimuli. identified before the means for achieving them Several authors have observed that strategic are evaluated. In a centralized structure thisdecisions, in addition to being a reaction to stim- means/ends relationship is closely related to theuli (problem and opportunity) can be proactively issue described above. Specifically, the goals ofinitiated by the interests of coalition members. coalition members will guide strategic decisionFor example, the great "leaps" of Mintzbergs making and they will have more impact on the(1973) "entrepreneurial" mode represent the pro- strategic process than under other structural con-active interests of a leader/founder who domi- ditions that are discussed later. Coalition mem-nates a small firm. A similar phenomenon has bers will also exhibit strong commitment to thebeen observed in larger organizations by Carter goals because they are their own (Latham &(1971) and Gerwin and Tuggle (1978). The lat- Yukl, 1975), even though they may be stated aster authors suggest that the strategic process can simply "to survive" or "be number one." Because 285 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  8. 8. of this fact, goals are likely to persist in the face process. In addition, experiments like those ofof significant changes in means, and the organi- Bruner, Goodenough, and Austin (1956) supportzation may continue to pursue strategic goals theoretical arguments regarding individuals ten-that have been rendered obsolete by changes in dency to simplify decision situations (Braybrookethe means of competition. & Lindblom, 1970; Schwenk, 1984). Therefore, in Itis widely acknowledged that a rational expla- an organization with a centralized structure, thenation of action, where an alternative is "chosen" cognitive limits of those few who dominate willbased on its ability to achieve a desired goal, determine how comprehensive the organizationdoes not capture the reality of organizational is in making strategic decisions.decision making. Instead of goal "maximizing" it The final issue regarding centralization con-is characterized as a process of "satisficing," and cerns its effect on an organizations ability to com-instead of being "rational" it is "intendedly prehensively integrate decisions to form an over-rational" (March & Simon, 1958). However, even all strategy. Mintzberg (1979) points out thatthough the decision process may not mirror a elaborate structural configurations (e.g., matrixrational model, a centralized structure gives top- organizations) are sometimes used to ensure thatlevel management an opportunity to make con- strategic decisions are integrated, but he arguesscious choices. Those choices are based on the that centralization offers a better solution. Withpreferences of an individual or small group, but such a structure, the comprehensiveness of inte-they are directly reflected in the organizations gration is still subject to the cognitive limitationsmoves. Therefore, structural centralization facili- of the dominant managers, but it may be highertates "strategic choice" (Child, 1972)and provides than if more formalized mechanisms are relieda modified (i.e., intendedly) "rational actor" on. For example, Quinn (1980) has suggested(Allison, 1971) explanation of strategic action. In that strategic planning systems are helpful, butaddition, because centralization makes it easier that integration is generally accomplished onlyfor those who dominate to pursue any proactive, in the mind of the top executive. Mintzberg (1978)opportunity-seeking interests that they may have, has provided support for this observation byit increases the likelihood that organizational arguing that a highly integrated "gestalt" strat-moves will be major (e.g., Mintzbergs leaps), as egy is likely only when the organization is con-opposed to incremental, departures from the trolled by a powerful leader. Therefore, whileexisting strategy. cognitive limits may restrict comprehensiveness It was suggested in the introduction that a in a centralized structure, they may have lessdecentralized structure accommodates members impact on the process of integration.cognitive limitations by factoring decision pro-cess responsibilities (March & Simon, 1958; Impact of FormalizationThompson, 1967). In contrast, a centralized struc- Structural formalization is characterized by theture is uniquely susceptible to those limitations, presence of rules and procedures that influenceand they affect how comprehensive the organi- decision-making behavior. As mentioned earlier,zation is in making strategic decisions. Many con- even when it exists only at low and intermediatetributions that question the ability of organiza- levels, formalization can affect an organizationstions to make decisions comprehensively are strategic process as participants gather and pro-grounded in well-documented work with indivi- cess information that is passed up the hierarchyduals. For example, Steinbruner (1974) has ar- (Carter, 1971). Moreover, Mintzberg (1979) hasgued that the characteristics of his "cybernetic" observed that firms that are highly formalized inmodel of organizational decision making are con- their "operating core" tend to be more formal-sistent with numerous cognitive theory princi- ized at all levels. In addition, the presence ofples that describe a noncomprehensive decision upper-level mechanisms such as budgeting sys- 286 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  9. 9. tems are known to have an impact on the strate- characteristic of interest). For example, formaliza-gic process (Bower, 1970), and strategic plan- tion can produce what Merton (1940) has charac-ning systems are a clear attempt to formalize de- terized as the "bureaucratic personality." In thiscision making at even the highest levels. The se- instance prescribed behaviors become ends incond column of Table 2 suggests that the impact themselves, and means become more importantof formalization can indeed be far-reaching. than ends. This phenomenon helps explain why Regarding issues of process initiation, it ap- formalized strategic planning processes some-pears that formalization increases the likelihood times degenerate into exercises that produce lit-that the strategic process will be motivated by tle more than a bound document. However,reactive (e.g., solving problems or crises), as Quinn (1980) suggests that the problem cannotopposed to proactive (e.g., searching for oppor- be overcome by simply formalizing strategic-leveltunities), behavior. For example, Steinbruner goals, because doing so activates organizational(1974) has characterized organizational decision processes that are difficult to reverse.making as a "servo-mechanism" whose pro- Most strategic management literature explainsgrammed responses are activated only when action as being the result of a conscious choice.critical variables get outside some specified However, discussion in this section has pointedrange. As is the case with Cyert and Marchs out that an organization that has a formalized(1963) description of problems triggering stan- structure is likely to respond to decision stimulidard operating procedures, decision stimuli may by employing standardized procedures (Cyert &be ignored if they are not monitored by the for- March, 1963; Steinbruner, 1974). The variablesmal system. In addition, it has been argued that that trigger the process are predetermined, andstrategic planning systems can become so for- so are the possible responses. Therefore, strate-malized that they drive out creative, proactive gic action in an organization with a formalizedbehavior (Lenz & Lyles, 1983; White, Dittrich, & structure is most accurately characterized byLang, 1980). This suggests that a formalized struc- Allisons (1971) "organization process" model. Itture has the inherent ability to discourage the is the "outcome" of a limited cadre of capabilities.pursuit of opportunities. In addition, the actions themselves are likely to A high level of formalization also affects the be incremental. Quinn (1980) has pointed out thatrole that goals play in the strategic process. For formalized strategic planning processes tend toexample, by prescribing bounds of behavior, for- institutionalize incrementalism because they pro-malized bureaucracies reduce goal incongrui- duce actions that are only marginal departuresties among members and provide reasonably from the existing state. Similarly, the presence ofwell-defined expectations about performance formal monitoring mechanisms encourages suchevaluation (Ouchi, 1978). Because of this fact, it organizations to make incremental adjustmentsis expected that strategic decisions will be made in response to feedback.with precise, as opposed to general, goals in As implied above, the degree of structural for-mind and that efficiency criteria will dominate malization will also affect how comprehensive(Simon, 1976). However, based on the above an organization is in making individual strategicargument regarding the reactive nature of the decisions. Rules and procedures contribute tostrategic process, under conditions of formaliza- the development of a firms repertoire of behav-tion these goals are more likely to be "remedial" iors, and they dictate how various decision-corrections of the present state, and not "posi- making activities will be handled. For example,tive," future intended states. formalized search procedures increase the likeli- The increased goal awareness that is brought hood that information will be sought from areason by formalization ultimately affects the rela- previously utilized, and that solutions that weretionship between means and ends (i.e., the third successful in the past will be used again (Cyert 287 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  10. 10. & March, 1963). Similarly, Carter (1971) has complexity do not become clear until each strate-described how strategic alternatives are evalu- gic process characteristic is considered indivi-ated by applying "threshold level" analysis that dually. (These effects are summarized in the finalallows projects to be automatically accepted or column of Table 2.)rejected if they are above or below certain levels The impact of complexity first becomes appar-on a specified variable. Therefore, although ent in its effect on how and where the strategicformalized, "planned behavior" is instituted to decision process is initiated. Specifically, becauseachieve rationality in decision making (Simon, an organizations structure imposes boundaries1976), the comprehensiveness of an organiza- of rationality on its members (Thompson, 1967),tions strategic process will be determined by the the degree of complexity specifies how wide orspectrum of behaviors that are accounted for in how narrow those boundaries will be (i.e., aits rules and procedures (i.e., the level of detail highly complex structure has many, narrowlyachieved). bounded positions). Therefore, since decision In addition to affecting how comprehensive or- stimuli (e.g., problems, opportunities) may makeganizations are in making individual strategic their initial appearance at any location in thedecisions, evidence suggests that formalization organization, the cognitive and motivational ori-can also affect how well those decisions are entation that is induced by a particular structureintegrated. For example, Schendel and Hofer will affect how a stimulus is perceived and acted(1979) point out that formal policies have been upon (Simon, 1974). Similarly, since strategicreplaced by formal planning systems as the pri- issues can emerge from anywhere in an organi-mary tool for trying to ensure that strategic deci- zation (Mintzberg, 1979), the degree of complex-sions are comprehensively integrated. Although ity will be a major determinant of whether mem-such systems may offer numerous benefits, their bers who are initially exposed to those issuesability to achieve a high level of integration is recognize them as having strategic significance,again determined by the detail of their design. or ignore them because of parochial perceptions.Moreover, even the most elaborate planning sys- Therefore, if structurally imposed bounds aretem may not be able to achieve comprehensive narrow, as they are with a high level of com-integration because ". . strategic decisions do plexity, members self-interests may lead themnot lend themselves to aggregation into a single to take no action, thereby leaving critical prob-massive decision matrix where all factors are lems and opportunities unattended or unex-treated simultaneously in order to arrive at a ploited.holistic optimum" (Quinn, 1978, p. 17). It is suggested that structural complexity also has an impact on the role that goals play in theImpact of Complexity strategic decision process. For example, Law- An organizations structure offers three poten- rence and Lorsch (1967) reported that a high leveltial sources of complexity: horizontal and verti- of complexity resulted in different goal orienta-cal differentiation, and spatial dispersion. How- tions across departments. This may explainever, the present description considers only Bowers (1970) observation that different peoplehorizontal and vertical differentiation because involved in the strategic process are motivatedthey best illustrate the dilemma that structure by different preferences. Therefore, in a com-poses as organizations try to accommodate plex organization the broad array of membersmembers cognitive limitations. Specifically, preferences or goals does indeed become aincreased division of labor, which is manifest as series of "constraints" on the decision processincreased horizontal and vertical differentiation, (Simon, 1964), which makes it unlikely that strate-requires increased coordination (Galbraith, gic decision making can successfully achieve1973). However, the pervasive effects of such some specific future state. In addition, by restrict- 288 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  11. 11. ing areas of responsibility and interest, a high nization is in making individual strategic deci-level of complexity increases the salience of sions. For example, Cyert and March (1963) haveindividuals goals, and makes it increasingly dif- argued that the "search" for decision-makingficult for organization-level goals to influence information is "biased" because participantsdecision making. selective perceptions cause them to focus on Regarding the relationship between means information that is salient to the interests of theirand ends, logic suggests that the level of com- department or unit. As mentioned above, indi-plexity will also have a strong influence on this vidual and unit biases are also introduced atcharacteristic. Task specialization accompanies multiple levels as information is "preprocessed"increased complexity, and specialization fosters on its way to the top (Carter, 1971). Therefore,parochial perceptions. Therefore, it is expected the comprehensiveness of the strategic decision-that members in an organization that has a com- making process will be affected by the extent toplex structure will have difficulty agreeing on which structural complexity evokes parochialgoals, and that the decision process will be itera- (either individual or work unit) behavior fromtive and political. Furthermore, because vast participants.goal differences may make it difficult to achieve The final issue regarding structural complex-consensus on ends, managers may have to be ity concerns the extent to which this dimensionsatisfied with obtaining agreement on means, affects an organizations ability to comprehen-even though they accept the means for different sively integrate decisions to form an overallreasons. It is therefore suggested that the multi- strategy. Although organizational strategy is usu-ple effects of structural complexity will combine ally characterized as a consciously integratedto produce strategic moves that are incremental, set of decisions, complexity creates problems forbut for reasons (i.e., goal differences) that are integration. As discussed earlier, a complexdifferent than was the case with a formalized structure assigns a restricted range of decisionstructure (i.e., institutionalization). process activities (e.g., information gathering or As implied above, conscious choice is also not analysis) to members in a variety of locationslikely to be an accurate explanation of strategic (i.e., departments and levels). While this ad-action in an organization with a complex struc- dresses members cognitive limitations, it alsoture. First, because information may have to pass increases the probability that actions taken inthrough multiple organizational levels, the out- one unit will not be consistent with those income of a strategic process can be affected another. Therefore, Mintzberg, (1978) has argued(Carter, 1971). More importantly, horizontal and that a highly integrated gestalt strategy will onlyvertical differentiation not only create differing be common early in an organizations life, whenpreferences among organizational members, structural complexity is low and power is cen-they also disperse power. These factors produce tralized.a constraint set that is not likely to be satisfied It is hoped that the dimension-specific proposi-and that must be attended to sequentially (Cyert tions presented in this section seem important& March, 1963). Therefore, as argued by Petti- and interesting. However, while these dimen-grew (1973), the division of labor that is manifest sions are frequently used in empirical research,in a complex structure explains strategic action practitioners and academics often think of struc-as the result of an internal political process, a ture in terms of different "types" (e.g., functionaldescription that is consistent with Allisons (1971) or divisional). Therefore, it could be both practi-"bureaucratic politics" model. cally and theoretically useful if the previous dis- As with the explanation of strategic action, the cussion could be extended to describe how theboundaries imposed by structural complexity also overall strategic process would look in organiza-have an impact on how comprehensive an orga- tions that have different types of structures. Such 289 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  12. 12. an extension would attempt to further illustrate tive structure to loosely link some combination ofthe strategic process/structure relationship and the other three. The "Adhocracy" is a large-scalepresent it in a context that has meaning for more matrix. Mintzberg also suggests that each struc-readers. Therefore, the arguments presented tural form results when one of several compet-above provide the basis for the next section, ing "pulls" dominates the others. If this argu-which discusses the "patterns" of strategic pro- ment is interpreted to mean that these structurescess characteristics that are likely to develop tend to be dominated by one dimension, there iswhen different types of structures are used. a basis for hypothesizing about the strategic pro- cess impact of different structural types. Patterns of Process Characteristics It is suggested that Mintzbergs (1979) purest forms-the Simple Structure, Machine Bureau-Alternative Structures and Pattems cracy, and Professional Bureaucracy-are in fact During the 1960s and 70s, several authors structures whose dominant dimension is one of(Burns & Stalker, 1961;Chandler, 1962;Lawrence the three that were previously discussed. More& Lorsch, 1967; Pugh, Hickson & Hinings, 1969; specifically, centralization is the dominant dimen-Rumelt, 1974) described alternative types of sion in a Simple Structure, formalization domi-structures. As a result of their contributions, inves- nates in a Machine Bureaucracy, while a Profes-tigators routinely refer to different structures with sional Bureaucracy is characterized first andnames such as "organic" and "mechanistic," foremost by complexity. This link is illustrated in"functional" and "divisional," or "workflow bur- Figure 1, where the three types are mappedeaucracy," with the expectation that readers will against the dimensions of structure. It is impor-have a basic understanding of their characteris- tant to note that these structures are widely distri-tics and implications. Therefore, it would be help- buted across the matrix, which emphasizes thatful if a link could be established between some they are very different. Also, it is later illustratedof these well-known forms and the strategic pro- that these types are the most common amongcess characteristics that were discussed in the organizations.previous section. For example, how and where is The relationship that appears to exist betweenthe process initiated? What role do goals play, Mintzbergs (1979) three types and the threeand so on, in an organization with an organic dimensions of structure suggests that the aggre-structure? How do these characteristics differ in gate of propositions that were previously attrib-an organization with a mechanistic structure? uted to each dimension can be used to produceThe empirical "archetypes" generated by Miller a pattern of characteristics that describe how stra-and Friesen (1977) represent one of the few at- tegic decisions are made in each type. In thetempts that have been made to establish a link final section each structural type is described, itsbetween structure and distinct patterns of strate- pattern of strategic process characteristics isgic process characteristics, but they differen- discussed, and the context where it can betiated structures primarily on the centralization expected to be most successful is characterized.dimension. It is important to note that the discussion of pat- In addition to the structural types described by terns draws heavily on the arguments that werethe above authors, Mintzbergs (1979) synthesis developed in the previous section. Therefore,of previous research (including those types) pro- only the primary conclusions (and not the sup-duced five forms that have begun to appear in porting literature) are provided.the literature. The "Simple Structure," "MachineBureaucracy," and "Professional Bureaucracy" Centralization and Simple Structureare the "purest" forms, while the "Divisional Mintzberg (1979) has argued that the SimpleForm" simply uses an over-arching administra- Structure is best characterized by what it is not. 290 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  13. 13. Centralization Centralized Decentralized Mchine Formalized ureaucracyFormalization l_______t Simple Professional Informal Structure Bureaucracy Simple Complex Simple Complex ComplexityFigure 1. Relative dimensions of three structural types.Specifically, it has little or no technical or ad- edge are concentrated, only the CEO can initi-ministrative support staff, little differentiation ate a response to problems or opportunities. Thebetween units, a "loose" division of labor, and a dominance of the CEO is also reflected in thevery small managerial hierarchy. In addition, explanation of strategic action-organizational"little of its behavior is formalized, and it makes actions reflect his or her intendedly rationalminimal use of planning, training, and. . .liaison choices. Similarly, with such a structure thedevices" (p. 306). The Simple Structure is a form actions that are chosen may be motivated by thewhere all important decisions are centralized in proactive, personal interests of the CEO, andthe hands of a dominant executive (CEO), who they will be made to achieve a "positive," thoughinformally coordinates the organizations func- general, goal. This argument illustrates Mintz-tional units. It is clearly a structure that is high in bergs (1979) observation that a potential benefitcentralization and low in both formalization and of the simple structure is its "sense of mission."complexity. Therefore, it is similar to Pugh et al.s However, because the goal reflects the CEOs(1969) "implicitly structured organizations." personal preferences, the organization may con- Since the Simple Structure is dominated by tinue to pursue it long after it should have beencentralization, its pattern of strategic process abandoned.characteristics can be predicted from the propo- An organization with a Simple Structure is notsitions provided earlier in the first column of restricted by formalized procedures or forced toTable 2. For example, because power and knowl- bargain among members who have different 291 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  14. 14. preferences, so when it takes strategic actions tionally grouped units at its lower levels, as wellthey are more likely to be major departures from as an elaborate administrative staff. Althoughits existing strategy. Moreover, the success or the work of those at the lowest levels may be thefailure of those actions can be directly attributed most directly controlled by formalization, it isto the CEO because his or her cognitive limita- important to recognize that ". . .at every hierar-tions are the primary constraint on the compre- chical level, behavior in a Machine Bureaucracyhensiveness of the strategic decision process. In is relatively more formalized than in other struc-addition, the CEO dominates the entire decision tural configurations" (Mintzberg, 1979, p. 318).process, and may well have an intimate knowl- The propositions previously presented in theedge of the firms daily operations. Therefore, it second column of Table 2 capture the pattern ofis more likely that decisions will be consistent strategic process characteristics that a Machineand integrated. Bureaucracy is likely to exhibit. For example, The Simple Structure is most successful in an because this structure is above all else formal-environment that is, using Duncans (1972) di- ized, it is likely that the process will be initiatedmensions, both simple (i.e., it has relatively few only when the condition of some formally moni-critical variables) and dynamic (i.e., those vari- tored variable indicates a need for action. Inables are shifting). Its pattern of strategic pro- addition, the strategic action ultimately taken willcess characteristics helps explain why. First, the reflect the application of one standardized re-CEOs dominance makes the organization di- sponse from among those that the organizationrectly dependent on his or her preferences and has developed. In combination, the above obser-cognitive capabilities, but the CEO has a realis- vations also suggest that the action will be takentic chance to understand a simple environment. to achieve a precise goal (e.g., a specified growthMoreover, the high level of centralization equips or profitability level), but that the goal will bethe CEO with an understanding of both operat- remedial (i.e., a correction to the initially moni-ing and strategic-level issues, which when com- tored deviation). However, because membersbined with the need for only one individual to recognize that their decision-making behavior isdecide, enables the organization to move quickly supposed to conform to specified rules and pro-when faced with environmental change. There- cedures, there is an increased likelihood thatfore, strategic decisions in a Simple Structure means will displace ends in a Machine Bureau-tend to be made quickly, in pursuit of positive cracy.opportunities, and with a sense of direction and As pointed out above, the dominating influ-integration. However, a complete dependence ence of formalization in this structure explainson the CEO is a constant source of risk with this strategic action as the output of standardizedtype of structure. organizational processes. These institutionalized processes have the added effect of producingFormalization and Machine Bureaucracy strategic actions that are only incremental depar- A Machine Bureaucracy is a structure that tures from the existing state. Moreover, the appro-relies on the standardization of work, which priateness of the actions taken in a Machinemakes it similar to the structures previously Bureaucracy is constrained, not by the CEOsdescribed by Inkson, Pugh, and Hickson (1970) cognitive limits, but by the level of detail that isand Pugh et al. (1969). Its most distinguishing achieved in its many systems (e.g., planning,features include ". . . very formalized procedures information) and processes. However, becausein the operating core, a proliferation of rules, such systems rely almost entirely on aggregated,regulations and formalized communication quantitative data that must be passed throughthroughout. . ." (Mintzberg, 1979, p. 315). In ad- multiple levels, they can be expected to yielddition, this structure tends to have large, func- only a moderately integrated strategy. 292 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  15. 15. A Machine Bureaucratic structure has its great- shown by Grinyer and Yasai-Ardekani, 1980, toest success in a simple, stable environment; the be related.)characteristics of its strategic decision process Since complexity is the dominant dimension ofsuggest why. An organization using this struc- a Professional Bureaucracy, the propositions pre-ture responds to formally monitored variables sented earlier in the final column in Table 2 high-and applies established standards in perform- light its pattern of strategic process characteris-ing its work. Therefore, the environment must be tics. Specifically, strategic problems or opportu-simple enough to allow critical variables to be nities may go unrecognized or ignored becauseidentified, and stable enough so that they can members interests are highly specialized, andbe tracked and standards developed. However, their perceptions parochial. This high level ofif the environment changes, even the highest horizontal specialization also increases the likeli-levels of the organization may be unresponsive hood that strategic action will be taken only afterbecause the need for change may be masked by extensive political bargaining among members,a dependence on information systems that gather or as the result of individual members applyingdata on a restricted range of variables. Moreover, solutions from their collective "garbage can" ofthere is an expectation among organizational skills (Cohen, March, & Olsen, 1972). In addition,members that strategic decision making should diversity among members and the salience ofproceed through formally established channels, their personal goals is likely to decrease thewhich may contribute to costly delays in a chang- impact of organization-level goals and produceing environment. strategic actions that are only incremental depar- tures from the organizations current state. Simi-Complexity and Professional Bureaucracy larly, parochial perceptions are the primary con- The "Professional Bureaucracy" is the name straint on the comprehensiveness of the strate-that Mintzberg (1979) gave to the structure most gic decision process, and they contribute to mak-frequently used in organizations such as gen- ing the integration of decisions quite low. Theseeral hospitals, universities, school systems, and observations are reflected in Mintzbergs (1979)social service agencies. These organizations rely conclusion that ". . . the notion of a strategy-aon highly trained professionals who control their single, integrated pattern of decisions commonown work, so the structure can accurately be to the entire organization-loses a good deal ofdescribed as very decentralized. Similarly, be- its meaning in a Professional Bureaucracy" (p.cause the work requires detailed knowledge of 363).specialized topic areas, the resulting structure is The work performed by Professional Bureau-horizontally complex and differentiated; vertical cracies is typically difficult to learn, yet quite welldifferentiation is limited. It should be emphasized defined (e.g., even complex surgical proceduresthat Professional Bureaucracies require stan- use widely agreed-upon techniques). Therefore,dardized behavior from their members, but that the environment is accurately described as com-behavior is achieved much differently than in a plex and stable. It is complex because it requiresMachine Bureaucracy. Members in this third skills learned only through advanced training,structural type are expected to enter with skills and stable because the necessary skills areand behavior standards established by their enduring enough to allow the profession toprofessions. However, these standards are not develop performance standards. In addition, thejust another type of formalization; they reflect a strength and divergence of members goalsseparate, independent dimension-complexity/ make such organizations highly political. There-specialization (Reimann, 1973). (The terms com- fore, the only apparent way that the executive-plexity and specialization are often used inter- level management of a Professional Bureaucracychangeably in the literature, and they have been can develop an overall strategy is by "patching" 293 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  16. 16. together the disparate project and program pref- tional perspective sheds light on the same pro-erences of the professionals, or by allocating cess in a firm that is dominated by formalization.resources only to those that offer apparent syn- In contrast, the small group, with all its socio-ergies. political phenomena, is the basic unit of analy- sis for understanding the strategic process in an Concluding Observations organization whose dominant dimension is com- This paper takes only a "first step" to encour- plexity. age investigation and debate on the strategic This paper also encourages a broader debate process/structure question. It is suggested, by assuming an uncommon perspective on or- though, that many critical issues have been ganizational structure. Most strategic manage- raised, and some answers have been offered. ment scholars continue to see structure as sim- For example, firms have been observed which ply a tool for implementing strategy, while orga- exhibit consistent "patterns of behavior" in mak- nization theorists discuss the relative effects that ing strategic decisions (Fredrickson & Mitchell, environment, technology, or size have on struc- 1984).The arguments presented here suggest that ture. It is suggested that each of these views is structures pervasive impact offers a reasonable unbalanced in its portrayal of structure. The explanation of why a firm develops a particular arguments presented here emphasize that a bal- way of making strategic decisions. More impor- anced view of the strategy/structure relationship tantly, these same arguments also suggest that must acknowledge that the strategic decision pro- alternative models of strategic decision making cess and its outcomes can be facilitated, con- (Allison, 1971; Mintzberg, 1973) are more than strained, or simply shaped by structures direct just different perspectives on the same phenom- effects. In accepting this argument, investigators enon. Organizations that differ in their dominant are not being asked to reject evidence that led tostructure are likely to make strategic decisions the "structure follows strategy" proposition. They using a very different process. are asked to recognize that there is a sizable The arguments presented here also provide a body of contributions which argues that andescription that is richer than those previously organizations structure may have importantavailable to explain why different structures have deterministic effects of its own. These effects havebeen associated with varying levels of perfor- neither been widely recognized nor investigatedmance in different contexts (Bums & Stalker, 1961; because the literature has been extremely frag-Lawrence & Lorsch, 1967; Khandwalla, 1977; mented and underdeveloped.Woodward, 1965). For example, it is the combin- While this paper may have raised some criti-ation of a dominant decision maker, who is pur- cal issues and offers a few tentative answers, itsuing positive goals, is willing to make major is only a first step. For example, the propositionsdepartures, has detailed knowledge of the entire presented here describe the strategic decisionorganization, and is faced with a situation that process in organizations where one of three struc-can be understood, that enables some small tural dimensions dominates. There may be cir-organizations to succeed in a rapidly changing cumstances where these dimensions interact toindustry in spite of being constrained by the deci- produce a strategic process whose pattern ofsion makers cognitive limitations. This example characteristics was not described. The empiricalalso sheds light on the recurring "unit of analysis" questions: when does structure follow strategy,issue that was most recently summarized by and when does structure, through its direct effectPfeffer (1982). More specifically, a firm that is on the strategic decision process, determinehighly centralized is likely to have a strategic strategy, remain. It is suggested that structure isdecision process that is best understood by using most likely to dominate in organizations wherean individual unit of analysis, while an organiza- an overall strategy is not institutionalized (i.e., 294 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  17. 17. either no explicit strategy has been articulated and test the primary arguments and ideas pre-or it is in the process of changing). Similarly, sented here, but that they will also refine andthere may be contexts such as crisis situations extend them. Such efforts are critical to under-where the effects of structure are subservient to standing the effects of structure and to develop-variables such as environment. Therefore, it is ing a balanced view of the strategy/structurehoped that investigators will not only challenge relationship. ReferencesAllison, G. T. (1971)Essence of decision: Explaining the Cuban Crozier, M. (1964) The bureaucratic phenomenon. English missile crisis. Boston: Little, Brown. translation, University of Chicago Press.Blackburn, R. S. (1982) Dimensions of structure: A review Cyert, R. M., & March, J. G. (1963) A behavioral theory of the and reappraisal. Academy of Management Review, 7, firm. Englewood Cliffs, NJ: Prentice-Hall. 59-66. Delbecq, A., Van de Ven, A. H., & Gustafson D. H. (1975)GroupBobbitt, H. R., & Ford, J. D. (1980) Decision maker choice as a techniques for program planning: A guide to Nominal determinant of organization structure. Academy of Man- Group and Delphi processes. New York: Scott, Foresman. agement Review, 5, 13-24. Duncan, R. (1972) Characteristics of organizational environ-Bourgeois, L. J., & Astley, W. G. (1979) A strategic model of ments and perceived environmental uncertainty. Ad- organizational conduct and performance. International ministrative Science Quarterly, 17, 313-327. Studies of Management and Organization, 6(3), 40-66. Duncan, R. (1979) What is the right structure? Decision treeBower, J. L. (1970) Managing the resource allocation pro- analysis provides the answer. Organizational Dynamics, cess. Cambridge, MA: Harvard University Press. 7, 59-80.Braybrooke, D., & Lindblom, C. E. (1970) A strategy of Fahey, L. (1981) On strategic management decision pro- decision: Policy evaluation as a social process. New York: cesses. Strategic Management Journal, 2, 43-60. Free Press. Fouraker, L. E., & Stopford, J. M. (1968) Organizational struc-Bruner, J. S., Goodenough, J. J., & Austin, G. A. (1956) A ture and the multi-national strategy. Administrative Sci- study of thinking. New York: Wiley. ence Quarterly, 13, 47-64.Burgelman, R. A. (1983) A model of the interaction of strate- Ford, J. D., & Slocum, J. W. (1977) Environment, technology gic behavior, corporate context, and the concept of stra- and the structure of organizations. Academy of Manage- tegy. Academy of Management Review, 8, 61-70. ment Review, 2, 561-575.Biums, T., & Stalker, G. M. (1961) The management of innova- Fredrickson, J. W. (1983) Strategic process research: Ques- tion. London: Tavistock. tions and recommendations. Academy of ManagementCarter, E. E. (1971) The behavioral theory of the firm and top Review, 8, 565-575. level corporate decisions. Administrative Science Quar- Fredrickson,J. W., & Mitchell,T. R. (1984)Strategicdecision terly, 16, 413-428. processes: Comprehensiveness and performance in anChandler, A. D. (1962) Strategy and structure: Chapters in industry with an unstable environment. Academy of Man- the history of the American industrial enterprise. Cam- agement Journal, 27, 399-423. bridge: MIT Press. Fry, L. W. (1982) Technology-structure research: Three criti- cal issues. Academy of ManagementJournal, 25, 532- 551.Child, J. (1972) Organizational structure, environment, and performance: The role of strategic choice. Sociology, 6, Fry, L. W., & Slocum, J. W. (1984) Technology, structure, and 1-22. workgroup effectiveness: A test of a contingency model.Child, J. (1974) Comments on Reimann and Mansfields Academy of Management Journal, 27, 221-246. "bureaucracy." Administrative Science Quarterly, 19, Galbraith, J. R. (1973) Designing complex organizations. 247-250. Reading, MA: Addison-Wesley.Cohen, M. D., March, J. G., & Olson, I. P. (1972) A garbage Gerwin, D., & Tuggle, F. D. (1978) Modeling organizational can model of organizational choice. Administrative Sci- decisions using the human problem solving paradigm. ence Quarterly, 17, 1-25. Academy of Management Review, 3, 762-773. 295 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  18. 18. Grinyer, P. H., & Yasai-Ardekani, M. (1980) Dimensions of Mintzberg, H., Raisinghani, D., & Theoret, A. (1976)The struc- organizational structure: A critical replication. Academy ture of unstructured decision processes. Administrative Sci- of Management Journal, 23, 405-421. ence Quarterly, 21, 246-275.Hage, J., & Aiken, M. (1967) Program change and organiza- Ouchi, W. G. (1978)The transmission of control through orga- tional properties. American Journal of Sociology, 72, nization hierarchy. Academy of Management Journal, 21 503-518. 173- 192.Hage, J., & Aiken, M. (1969) Routing technology, social struc- Perrow, C. (1972) Complex organizations: A critical essay. ture and organizational goals. Administrative Science Glenview, IL: Scott, Foresman. Quarterly, 14, 368-379. Pettigrew, A. (1973) The politics of organizational decision-Hall, R. H. (1977)Organizations: Structure and process. Engle- making. London: Tavistock. wood Cliffs, NJ: Prentice-Hall. Pfeffer, J. (1982) Organizations and organization theory.Hedberg, B. L. T., Nystrom, P. C., & Starbuck, W. I. (1976) Boston: Pitman. Camping on seesaws: Prescriptions for a self-designing Pugh, D. S., Hickson, D. J., & Hinings, C. R. (1969) An empiri- organization. Administrative Science Quarterly, 21 41-65. cal taxonomy of structures of work organizations. Ad-Inkson, J. H. K., Pugh, D. S., & Hickson, D. I. (1970) Organi- ministrative Science Quarterly, 14, 115-126. zation, context and structure: An abbreviated replication. Pugh, D. S., Hickson, D. J., Hinings, C. R., & Turner, C. Administrative Science Quarterly, 15, 318-329. (1968) Dimensions of organization structure. AdministrativeJelinek, M. (1977) Technology, organizations and contingency Science Quarterly, 13, 65-91. theory. Academy of Management Review, 2, 17-26. Quinn, J. B. (1978) Strategic change: Logical incrementalism.Khandwalla, P. N. (1977) The design of organizations. Har- Sloan Management Review, 20(1), 7-21. court Brace Jovanovich. Quinn, J. B. (1980) Strategies for change: Logical incremen-Latham, G. P., & Yukl, G. A. (1975) A review of research on talism. Homewood, IL: Irwin. the application of goal setting in organizations. Academy of Management Journal, 18, 824-845. Reimann, B. C. (1973) On the dimensions of bureaucratic structure: An empirical reappraisal. Administrative Sci-Lawrence, P. R., & Lorsch, J. (1967) Organization and envi- ence Quarterly, 18, 462-476. ronment. Boston: Harvard Business School. Rumelt, R. P. (1974) Strategy, structure and economic perfor-Lenz, R. T., & Lyles, M. A. (1983) Crippling effects of "hyper- mance. Cambridge, MA: Harvard University Press. rational" planning. College of Commerce and Business Administration, University of Illinois at Urbana-Cham- Schendel, D. E., & Hofer, C. W. (Eds.). (1979) Strategic man- paign, Faculty Working Paper No. 956. agement. Boston: Little, Brown.Lindblom, C. E. (1959) The science of muddling through. Schwenk, C. R. (1984) Cognitive simplification processes in Public Administration Review, 19, 120-128. strategic decision-making. Strategic Management Journal. 5, 111-128.March, J. G., & Simon, H. A. (1958) Organizations. New York: Wiley. Simon, H. A. (1964) On the concept of organizational goal. Administrative Science Quarterly, 9, 1-22.Merton, R. K. (1940) Bureaucratic structure and personality. Social Forces, 18(4), 560-568. Simon, H. A. (1974) How big is a chunk? Science, 183(4124), 482-488.Miller, D., & Friesen, D. H. (1977) Strategy-making in con- text: Ten empirical archetypes. Journal of Management Simon, H. A. (1976) Administrative behavior (3rd ed.). New Studies, 14, 253-280. York: Free Press.Mintzberg, H. (1973) Strategy making in three modes. Cali- Steinbruner, J. D. (1974) The cybernetic theory of decision. fornia Management Review, 16(2), 44-53. Princeton, NJ: Princeton University Press.Mintzberg, H. (1978) Patterns in strategy formation. Manage- Thompson, J. D. (1967) Organizations in action. New York: ment Science, 24, 934-948. McGraw-Hill.Mintzberg, H. (1979) The structuring of organizations. Engle- Van de Ven, A. H. (1976) A framework for organization wood Cliffs, NJ: Prentice-Hall. assessment. Academy of Management Review, 1, 64-78. 296 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions
  19. 19. Walton, E. J. (1981) The comparison of measures of organiza- complexity on implementation attempts. Administrative Sci- tion structure. Academy of Management Journal, 6, 155- ence Quarterly, 25, 428-440. 160.White, S., Dittrich, J. E., & Lang, J. R. (1980) The effects of Woodward, J. (1965) Industrial organizations: Theory and group decision making process and problem situation- practice. Oxford University Press. James W. Fredrickson is Associate Professor of Business in the Graduate School of Business, Columbia University. 297 This content downloaded on Sun, 30 Dec 2012 15:04:12 PM All use subject to JSTOR Terms and Conditions

×