Organizations do not always adopt the structures called for by their imper- sonal conditions—their ages and sizes, the technical systems they use, the stability, complexity, diversity, and hostility of their environments. A number of power factors also enter into the design of structure, notably the presence of external control of the organization, the personal needs of its various members, and the fashion of the day, embedded in the culture in which the organization finds itself (in effect, the power of social norms). Three hypotheses describe a number of the findings about these power factors:
Hypothesis 14: The greater the external control of the organization, the more centralized and formalized its structure. A number of studies of both public and private organizations have provided evidence that outside control of them—whether directly by specific owners or indirectly, say, by a major supplier on whom they are dependent—tends to concentrate their decision-making powers at the top of their hierarchies and to encourage greater than usual reliance on rules and regulations for internal control. All this, in fact, seems logical enough. The two most effective means to control an organization from the outside are (1) to hold its most powerful decision maker—its chief executive officer—responsible for its actions, and (2) to impose clearly defined standards on it, transformed into rules and regula- tions. The first centralizes the structure; the second formalizes it.
Moreover, external control forces the organization to be especially careful about its actions. Because it must justify its behaviors to outsiders, it tends to formalize them. Formal, written communication generates re- cords that can be produced when decisions are questioned. Rules ensure fair treatment to clients and employees alike. External control can also act to bureaucratize the structure by imposing on it more sweeping demands than usual for rationalization. For example, whereas the autonomous firm can deal with its suppliers and clients in the open market, the subsidiary may be informed by headquarters that it must purchase its supplies from a sister subsidiary, and moreover that managers of the two subsidiaries must sit down together to plan the transfers in advance so that no surplus or shortages will result. Or a parent organization or government might insist on standards being applied across the whole range of organizations it controls. It may demand anything from the use of a common logo, or corporate symbol, to a common management information system or set of purchasing regulations. Entrepreneurial firms with organic structures that are purchased by larger corporations are often forced to develop organi- grams, specify job descriptions and reporting relationships more clearly, and adopt action planning and a host of other systems that bureaucratize their structures.
To conclude, Hypothesis 14 indicates that when two organizations are the same age and size, use the same technical system, and operate in the same environment, the structure of the one with the greater amount of external control—by government, a parent organization, the unions, or whatever—will be more centralized and more formalized. This, of course, raises all kinds of interesting issues in societies that find more and more of their autonomous organizations being gobbled up by giant conglomera- tions—big business, big government, big labor. The loss of autonomy means not only the surrender of power to the external controller but also significant changes within the structure of the organization itself, no mat- ter what its intrinsic needs—more power concentrated at its strategic apex, tighter personnel procedures, more standardization of work processes, more formal communication, more regulated reporting, more planning and less adapting. In other words, centralization of power at the societal level leads to centralization of power at the organizational level, and to bureaucratization in the use of that power.
Hypothesis 15: The power needs of the members tend to generate structures that are excessively centralized. All members of the organiza- tion typically seek power—if not to control others, at least to control the decisions that affect their own work. The managers of the strategic apex promote centralization in both the vertical and horizontal dimensions; the managers of the middle line promote vertical decentralization, at least down to their own levels, and horizontal centralization to keep power within the line structure; the analysts and the support staff favor horizontal decentralization, to draw power away from the line managers; and the operators seek vertical and horizontal decentralization, all the way down to the operating core.
But the dice of this power game are loaded. To function effectively, organizations typically require hierarchical structures and some degree of formal control. And these naturally put power in the hands of the line managers, as opposed to the staff specialists or the operators, and aggre- gate that power at the top of the hierarchy, in the hands of the managers of the strategic apex. We have seen that various situational factors—such as a sophisticated technical system and environmental complexity—call for a sharing of central power. But to the extent that the line managers, notably the senior ones, relish power, the structure can easily become excessively centralized. That is, more power can be concentrated at its top than the factors of age, size, technical system, and environment would normally call for (at least until the resulting inefficiencies catch up with the organi- zation).
Hypothesis 16: Fashion favors the structure of the day (and of the culture), sometimes even when inappropriate. Stinchcombe’s research, discussed in Hypothesis 2, suggests that there is such a thing as “the structure of the day”—that is, the one favored by industries founded in a given period. But his research also shows that structures transcend peri- ods; in other words, that some organizations retain structures favored in previous periods. The implication of this is that when a new structure comes along, it is appropriate for some organizations but not for others.
This point has, apparently, been lost on a good many organizations, because fashion—the power of the norms of the culture in which the orga-nization finds itself—seems to play an important role in structural design. We might like to believe that organizations are influenced only by factors such as age, size, technical system, and environment, not by what Jones, Inc., is doing next door. But there is too much evidence to the contrary. Part of the problem probably lies with the business periodicals and consulting firms eager to promote the latest fad. As Whistler (1975) has noted, “There is still money to be made, and notoriety to be gained, in peddling universal prescriptions. In economic terms, the demand is still there, in the form of executives who seek the gospel, the simple truth, the one best way” (1975:4). Paris has its salons of haute couture; likewise, New York has its offices of “haute structure,” the consulting firms that bring the latest in high structural fashion to their clients—long-range planning (LRP), management information systems (MIS), management by objectives (MBO), organization development (OD).
In the 1960s, the management media heralded “the coming death of bureaucracy,” to use the title of an article by Warren Bennis (1966). And many organizations took this seriously, some to their regret. Thus, when Lawrence and Lorsch describe the low-performance container firm that tried to use integrators—one of the very fashionable tools of organic struc- ture—in a simple, stable environment, we find fashion extracting its toll in inappropriate structural design. Since Bennis’s article, it has become evi- dent that bureaucracies will not die. Not as long, at least, as organizations grow old and large, mass-produce their outputs, and find simple, stable environments to nurture their standards. The fact is that articles would not be published and speakers would not attend conferences to tell of “the one best way” if the printers and airlines were not structured as bureaucracies. Today, few would deny that bureaucracies are alive, if not well.
Throughout this century, the swings between centralization and de- centralization at the top of large American corporations have resembled the movements of women’s hemlines. But the trend toward the use of divi- sionalization has been consistent, ever since du Pont and General Motors first made it fashionable in the 1920s. Thus, Rumelt found in a study of the Fortune 500 strong support not only for Chandler’s (1962) well-known proposition that “structure follows strategy” but for another, that “struc- ture also follows fashion” (1974:149). The use of the divisionalized form increased from 20 percent in 1949 to 76 percent in 1969; but not all of it was explained by market diversification, as Hypothesis 11 would have us be- lieve: “Until the early 1960s the adoption of product-division structures was strongly contingent upon the administrative pressures created by di- versification but .. . in more recent years divisionalization has become accepted as the norm and managements have sought reorganization along product-division lines in response to normative theory rather than actual administrative pressure” (p. 77).
Of course, fashionable structure need not be inappropriate structure.
Fashion reflects new advances in organizational design, advances that suit some organizations with older structures. Once the divisionalized form became established, it was appropriately adopted by most diversified com- panies that had been structured along functional lines.12 Indeed, those that failed to do so were saddled with structures that suddenly became out of date—less effective than the new alternative. Much like the dowager who always dresses as she did in her heyday, so too the organization may cling to a structure appropriate to days gone by. Thus, one study found that in the absence of competitive pressures, some European companies did not divisionalize even though they were diversified. Placid environments en- abled them to retain outdated, ineffective structures (Franko, 1974).
This finding also suggests that structural fashion is in some sense culture-bound. What is all the rage among the Fortune 500 (the largest U.S. corporations) may simply look odd to the Fortune 200 (the largest non-U.S. corporations). West Virginians and Westphalians may simply have different preferences for structure. This is another way of saying that culture, working through fashion, is another factor that influences struc- tural design.
The literature provides evidence for this too, for example, that certain European societies—such as the German—take better to bureaucracy than does the American, or that the Japanese place much heavier emphasis on indoctrination than do most other people.
In contemporary American culture, we see quite different trends in structural fashion. Coming quickly into vogue, close behind the division- alized form, is project structure, what Bennis and Slater (1964) and then Toffler (1970) have called “ad-hocracy”—in essence, selectively de- centralized organic structure that makes heavy use of the liaison devices. One can hardly pick up a management journal without reading about task forces, integrating managers, matrix structure. Clearly, this structure cor- responds well to the calls for the destruction of bureaucracy, to the demo- cratic norms prevalent in American society, and to its increasingly better- educated work force. But although this may be the structure of our age— well-suited to new, “future-shocked” industries such as aerospace and think-tank consulting—it may be wholly inappropriate for most older in- dustries. It, too, is no panacea. Like all the structures before it, themselves once fashionable, it suits some organizations and not others. It is to be hoped that those others will not opt for project structure, as did one of Lawrence and Lorsch’s container firms, just because it is fashionable.
To conclude our discussion of the situational factors, we note that different ones tend to affect the structure at different levels, although a number can affect the same design parameter (as in the case of formaliza- tion of behavior, which is affected by age, size, technical system, environ- mental stability, and culture). The factors of age and size, although signifi- cant at all levels, seem most pronounced in the middle of the structure; that is where, by creating changes in the favored mechanism of coordination, they produce extensive structural elaboration. The technical system, being housed in the operating core, clearly has its greatest effect there. But it has important selective effects elsewhere as well—for example, at middle levels requiring an extensive support staff when it is sophisticated. The environ- mental factors seem to have exactly the opposite effect from the technical- system ones. It is the managers and staff specialists at and near the strate- gic apex, those who must function continuously at the organization’s boundaries, who are most affected by the environmental dimensions. These dimensions also importantly affect the structure in the middle, but have only a selective effect on the operating core, which the rest of the structure in fact tries to seal off from direct environmental influence. Final- ly, the power factors seem to cut across all levels of the structure, but only on a selective basis. External control, member needs for power, fashion, and culture sometimes modify the structures that would otherwise result from consideration of only the factors of age, size, technical system, and environment.
Source: Mintzberg Henry (1992), Structure in Fives: Designing Effective Organizations, Pearson; 1st edition.