1. Market diversity
One situational factor above all drives the organization to use the Divi- sionalized Form—market diversity. The organization faced with a single integrated market simply cannot split itself into autonomous divisions; the one with distinct markets, however, has an incentive to create a unit to deal with each. This enables the organization to manage its strategic port- folio centrally, while giving each component of that portfolio the un- divided attention of one unit.
This relationship between diversification and divisionalization has received a good deal of support in the literature. Research has demon- strated the spread of structural divisionalization as a direct response to product diversification. This has proceeded in waves, particularly since World War II, encompassing most of the Fortune 500 and then jumping across the Atlantic to large corporations in England, Germany, France, and Italy.
In our discussion of the conditions surrounding the other configura- tions, we noted that not only do the situational factors influence the choice of the design parameters; the design parameters also influence the “choice” of the situational factors. In other words, the situational factors form a part of the configurations we are discussing. Here we have an excellent example of this. Chandler argued that structure follows strategy, that structural divisionalization results from strategic diversification. But the opposite relationship has held up in some research as well, that divi- sionalization encourages further diversification. The ease with which headquarters can add new divisions in this structure encourages it to do so; moreover, divisionalization generates a steady stream of general managers who look for more and larger divisions to run.
. . . the divisional structure becomes a built-in “school of management,” training middle level general managers in the problems and opportunities associated with economic responsibility. As a result, this form of organization provides a pool of trained talent from which to draw, a pool from which a new group may be formed in a few days or weeks to take over and manage a new activity. Both the structure and the internal “schooling” facilitate rapid and profitable exploitation of new ideas, a key element in the growth strategies of the [divisionalized] firms. (Scott, 1971:14)
This seems to explain why one study found that the more diversified American firms on the domestic scene were the ones most likely to develop international operations: when new products ran out at home, the aggres- sive young managers could be satisfied with foreign subsidiaries to manage.
In Chapter 3 we discussed three kinds of market diversity—product and service, client, and region. In theory, all three can lead to divisionaliza- tion. Physically dispersed markets, for example, create communication problems that give the organization an incentive to set up geographical divisions to deal with each region, as in retail chains, post offices, and railroads serving large areas. Add to this high transportation costs—as in the case of a cement manufacturer—and there is further incentive to divi- sionalize on a regional basis. Yet, based on client or regional diversifica- tion in the absence of product or service diversification, divisionalization often turns out to be incomplete. With identical products or services in each region or for each group of clients, the headquarters is encouraged to centralize a good deal of decision making and concentrate a good deal of support service at the center, to ensure common operating standards for all the divisions. This centralization and concentration of certain functions— some of them critical in formulating product-market strategies—seriously reduces divisional autonomy. In effect, the structure is driven toward inte- grated Machine Bureaucracy, but with one difference: Its operations are divided into distinct market-based units. Thus, one study found that insur- ance companies concentrate the critical function of investment, and re- tailers that of purchasing. The headquarters of the latter control sources of supply, product range, pricing, and volume terms, as well as site and property development and merchandising. Day-to-day operations of the retail stores are left to the store managers, who are supervised by a regional hierarchy.
We shall use the term carbon-copy bureaucracy for this hybrid of Divi- sionalized Form and Machine Bureaucracy, the structure that results when an organization sets up identical regional divisions and then concentrates certain critical functions at headquarters. Each division is a replica—a car- bon copy—of all the others, performing the same activities in the same ways, unique only in its location. The carbon-copy bureaucracy is, in fact, found in all the examples given above of regional divisionalization, but it is probably most common in retailing—the supermarket chain with fifty identical stores, the post office with a duplicated facility in each city of the nation, the motel or fast-food franchise, where, once inside, customers can hardly tell whether they are in Driggs, Idaho, or Dublin, Ireland.
The carbon-copy bureaucracy can also be found in the manufacturing sector, where a simple and stable environment and standard products drive the structure toward Machine Bureaucracy, but dispersed markets coupled with either high transportation costs or perishable products en- courage the organization to replicate its production facilities in different regions. Common examples are ba’keries, breweries, cement producers, and soft-drink bottlers. They produce and market their products in each city of any size, subject to tight standards set and enforced by the central headquarters. (The recent introduction of a small oven in our local bread store—part of a chain operating exclusively in the Montreal area—suggests that manufacturing carbon-copy bureaucracies can exist on small scales indeed.)
The giant multinational enterprise with identical product lines in vari- ous national markets also tends to resemble the carbon-copy bureaucracy. A division or “subsidiary” is created in each market to manufacture and distribute the products subject to the dictates of headquarters. In other words, certain critical functions—most notably product development—are retained by the central administration. Of course, the more foreign the subsidiary, the more it needs the power to adapt the products and market- ing techniques to its local conditions; in other words, the greater is the pull to pure divisionalization. But the multinational enterprise can avoid that pull by concentrating on products that can be standardized throughout the world (Coca-Cola being the classic example), and by avoiding very foreign markets. Thus, American corporations have typically expanded first into Canada—close, convenient, and minimally foreign—then into Europe, later perhaps beyond, but not frequently to the cultures most foreign to the West.
In Canada, in fact, the phenomenon of the headless subsidiary—one with no control over its main strategies—is so common that it has merited a special name: the miniature replica effect. It is set up in Canada to produce products designed in the United States according to American specifica- tions on production lines engineered by the American technostructure. It is interesting how often these firms have reacted to attacks by Canadian nationalists with the claim that all their employees but one are Canadian nationals. That one is, of course, the president, placed on the shoulder’s of the subsidiary to receive the orders from its brain in New York.
2. Technical system
What of the role of the other situational factors—besides market diver- sity—in the Divisionalized Form? In one sense, technical system is a factor, specifically its economies of scale. Divisionalization is possible only when the organization’s technical system can be efficiently separated into seg- ments, one for each division. For example, whereas a geographically di- versified cement company can duplicate its processing facilities many times across the face of the nation, a likewise diversified aluminum com- pany with the same sales volume may be unable to if it cannot afford more than one smelter. And so the aluminum company retains a functional structure. (Even for the cement producer, divisionalization may be in- complete: Geographical diversification, as noted above, tends to encourage a functional-divisional hybrid, often the carbon-copy bureaucracy.) When it is the product lines rather than the geographical regions that are diversi- fied, separation of the technical system usually takes place naturally, no matter what the economies of scale; different product lines require differ- ent technical systems to begin with.
There is, however, evidence of a more important, although indirect, relation between economies of scale and divisionalization. Organizations that must devote huge capital resources to very high fixed-cost technical systems—steel and aluminum producers, and other “heavies” of Ameri- can industry—tend not to diversify in the first place, and so not to division- alize. To be more precise, as a group they show little enthusiasm for “hori- zontal” diversification—into parallel or unrelated product lines. They do diversify “vertically,” moving into the product lines at the two ends of their production chains, thereby becoming their own suppliers and cus- tomers. But as we shall see later in this chapter, the strong interdepen- dences between product lines in the same production chain leads to an incomplete form of divisionalization.
3. Environment
In respect to the factors of environment, the Divisionalized Form differs funamentally from the other four configurations. Each of those has its own particular environment, specifically one of the four boxes of the static- dynamic, simple-complex matrix discussed in Chapter 6. In other words, whereas it is primarily the broad environmental dimensions of stability and complexity that position the other configurations, it is another, more restricted environmental dimension—market diversity, in particular, product diversity—that positions the Divisionalized Form. This narrows its range of application considerably compared with the other four configurations.
Nevertheless, the Divisionalized Form does have a preferred environ- ment, which it shares with the Machine Bureaucracy. That is because of another condition prerequisite to the use of the Divisionalized Form— outputs (specifically performance criteria) that can be standardized. As we saw in the last chapter, complex environments lead to outputs that cannot be measured or standardized. Likewise, in dynamic environments, out- puts and performance standards cannot easily be pinned down. So the Divisionalized Form works best in environments that are neither very complex nor very dynamic; in fact, the very same environments that favor the Machine Bureaucracy. This leads to a rather precise specification of the conditions that most commonly accompany this configuration: the Divi- sionalized Form is the structural response to a Machine Bureaucracy, operating in a simple, stable environment (typically without huge econo- mies of scale), that has diversified its product or service lines horizontally.
When an organization attempts to force divisionalization on units operating in other kinds of environments—complex or dynamic ones— where the outputs cannot be measured by performance controls, a hybrid structure normally results. In effect, the headquarters must rely on some mechanism other than the standardization of outputs to control the divi- sions. If it turns to rules and regulations—in effect, the imposition of stan- dards that control decisions and work processes of the divisions directly— then a hybrid results with Machine Bureaucracy, similar to the carbon-copy bureaucracy. If, instead, the headquarters managers increase their person- al surveillance (direct supervision) of the divisions through more frequent contact with their managers, then a hybrid with Simple Structure results, which we can call the personalized divisionalized form. Alternatively, should they seek to control the behavior of the divisions primarily through social- ization—in effect, appointing only managers they can trust fully because these have been through an extensive program of indoctrination or for some other reason identify strongly with it—then a hybrid with some characteristics of Professional Bureaucracy emerges, which we can call the socialized divisionalized form.
Competition is another variable that has been suggested as an en- vironmental determinant of the Divisionalized Form. In particular, Franko (1974) concluded in a study of European multinational firms that the ab- sence of competition may delay the adaption of the Divisionalized Form despite product diversification. He found that European companies oper-ating in cartels and the like tended to maintain their functional structures long after they diversified. Likewise, Scott (1973:141) found the most rapid spread of divisionalization in America during periods when competitive pressures were maintained by antitrust legislation and economic condi- tions, and in Europe when competitive pressures were generated by the Common Market and by supply catching up with demand in the 1960s. This argument makes sense, but it is not unique to the Divisionalized Form, it is the need for efficiency that drives all organizations to make sure their structures match their situation. (That was the point of the con- gruence hypothesis presented in Chapter 6.) Structural change always lags situational change, the length of that lag affected by the pressures to be efficient. Competitive pressures figure prominently among these, not only forcing a shift to the Divisionalized Form soon after product diversification, but also presumably forcing a quick shift back to the functional form should the organization later consolidate its product lines.
4. Age and size
What about the factors of age and size? Although large size itself does not bring on divisionalization, surely it is not coincidental that most of Amer- ica’s largest corporations use some form of this structure. The fact is that as organizations grow large, they become prone to diversify and then to divisionalize. One reason is protection: Large manufacturing firms tend to be organized as Machine Bureaucracies, structures that, as we noted in Chapter 9, try to avoid risks. Diversification spreads the risk. Also, the larger a firm becomes vis-a-vis its competitors, the more it comes to domi- nate its traditional market. Eventually, it simply runs out of room for expansion (because there is no market share left or because its dominance has come to the attention of the antitrust regulators), and so it must find further growth opportunities elsewhere. Thus it diversifies, and later must divisionalize. Moreover, as noted earlier, divisionalization creates a cadre of aggressive general managers who push for further diversification and further growth. So we must conclude that there is, in fact, an important relationship between size and divisionalization, with diversification the intermediate variable. The giant corporations—with the few exceptions that remain in one business because of enormously high fixed-cost techni- cal systems—not only require divisionalization but were able to reach their giant size only because of it.
In fact, many corporations have grown so large and diversified that the simple Divisionalized Form is not sufficient for them. They make use of a variant we call the multiple-divisionalized form, with divisions on top of divisions. For example, regional divisions may be superimposed on prod- uct divisions, or broad product divisions (“groups”) may be superimposed on narrower ones, as in the case of General Electric, shown later in this chapter in Figure 11-5.
Like size, age is also associated with the Divisionalized Form. In larger organizations, the management runs out of places to expand in the traditional markets; in older ones, the managers sometimes get bored with the traditional markets and find diversion through diversification. In other cases, time brings new competitors into old market niches, forcing the management to look for new ones with better potential. Thus, with divi- sionalization most common among the largest and oldest corporations, the Divisionalized Form emerged in Chapter 6 as the third stage of structural development, following Maching Bureaucracy.
The Divisionalized Form need not, however, always follow other configurations at a late stage of development. Some organizations, in fact, begin their lives with it. They divisionalize from without, so to speak; that is, they agglomerate rather than diversify. Independent organizations that join together to form new alliances—perhaps to benefit from the sharing of financial resources or support services—but are intent on guarding as much of their previous autonomy as possible naturally prefer a variant of the Divisionalized Form. These alliances, generally known as associations or federations, occur when farmers create cooperatives to market their pro- duce and when small construction firms do likewise to match the power of large unions or bigger competitors. Of course, not all agglomerations are voluntary: Stock-market operators take over corporations in proxy fights and force them into federations, as do governments when they nationalize firms to pool their resources for purposes of national planning or the devel- opment of the scale needed to meet foreign competition. When the units of the federated organization produce common products or services, strong pressures naturally arise to consolidate their activities into a tighter struc- ture—specifically to concentrate critical functions at the administrative headquarters—and the divisionalized structure tends to be driven to an integrated Machine Bureaucracy one.
5. Power
These last points introduce our final set of situational factors, those related to power, which also play a role in the Divisionalized Form configuration. We have just seen that power can explain federation: Small organizations need to band together to match the power of the bigger ones, and govern- ments or owners use their power to force unwilling partners to federate. We also saw earlier the role of power within the structure, that of the division managers who encourage growth, diversification, and divisional- ization to enhance their own positions. Even in the functionally structured organization, the drive by the aggressive middle manager for more autono- my amounts to a pull to divisionalize at his level of the hierarchy. And in the case of the top manager, the Divisionalized Form is by far the most effective structure by which to increase the power of his overall organiza- tion, since it enables units to be added with relatively little effort and disruption. (Internally, the top manager must, of course, share much of that increased power with the divisional managers.) Indeed, the waves of conglomerate diversification in U.S. industry appear to represent a giant power game, with corporate chief executives vying with each other to see who can build the largest empire.
These same factors of power have hardly been absent in other spheres as well, helping to explain the growth in popularity of the Divi- sionalized Form in unions, school systems, universities, and especially governments. Thus, we have the story of the president of a multiversity— one public university among six in a Canadian province—who justified his attempt to take over the two smallest ones with the argument that it would be more “convenient” for the government to negotiate with four admin- istrations instead of six. No mention of augmenting his power, no mention of the costs of his administration having to negotiate with two new cam- puses, no mention of the effects on those two small Professional Bureau- cracies of the introduction of another, intermediate layer of supervision.
As government grows larger—itself often spurred on by similar “con- venient” power grabs—it is forced more and more to revert to a kind of Divisionalized Form. That is, the central administrators, being unable to control all the agencies and departments (divisions) directly, settle for granting their managers considerable autonomy and then try to control their performance. One can, in fact, view the entire government as a giant Divisionalized Form (admittedly an oversimplification, since all kinds of interdependences exist among the departments), with its three main coor- dinative agencies corresponding to three main forms of control used by the headquarters of the divisionalized organization. The budgetary agency, technocratic in nature, concerns itself with performance control of the de- partments; the public service commission, also partly technocratic, con- cerns itself with the recruiting and training of government managers; and the executive (or Privy Council) office reviews the major proposals and initiatives of the departments. Perhaps this concept of the government as a giant Divisionalized Form is taken to its natural conclusion in the commu- nist state, where public corporations and other agencies are tightly regu- lated by planning and control systems operated by a powerful central technostructure.
Finally, there is fashion, not an insignificant factor in the popularity of the Divisionalized Form. Our comments above suggest that this struc- tural form is becoming increasingly popular in the public and institutional sectors. In the private sector, as noted, divisionalization became fashion- able after the restructuring of du Pont and General Motors in the 1920s. Since that time, American corporations have undergone a number of waves of such structural change. Much of this was, as we have seen, stimulated by diversification. But not all. As one student of the Fortune 500 noted in looking at his data on divisionalization, structure also follows fashion (Rumelt, 1974:149). In recent years, some managements have re- organized “in response to normative theory rather than actual administra- tive pressure” (p. 77). In Europe, until recently, the Divisionalized Form was unfashionable, with many diversified corporations resisting its use. Now the pendulum has swung the other way, and no doubt some corpora- tions with integrated markets have been carried along, to their eventual regret.
Source: Mintzberg Henry (1992), Structure in Fives: Designing Effective Organizations, Pearson; 1st edition.