It has been difficult up to this point to keep from discussing technology as a factor in organization design. Clearly, structure is tightly intertwined with it. But before considering how, we must make quite clear what we mean by the terms we shall use.
Technology is a broad term that has been used—and abused—in many contexts. We prefer to avoid it. For its broader meaning—essentially, the knowledge base of the organization—we shall use the term complexity and discuss it under environment. Here we shall focus on a narrower in- terpretation of technology—namely, the instruments used in the operating core to transform the inputs into outputs, which we shall call the technical system of the organization. Note that the two concepts are distinct. Accoun- tants, for example, apply a relatively complex technology (that is, base of knowledge), with a simple technical system—often no more than a sharp pencil. Alternatively, most people drive automobiles without ever know- ing what goes on under the hood; in other words, they use a fairly complex technical system with hardly any technological knowledge at all.
In discussing the effect of the technical system on the structural pa- rameters, we find it convenient to introduce our framework or organiza- tional types first, and then turn to hypotheses.
1. Woodward’s study of unit, mass, and process production
We have already referred to Joan Woodward’s pathbreaking analysis of the effects on structure of different forms of technical systems used in indus- try. Woodward focused on three basic systems of production—unit (essen-tially custom), mass (of many standard items), and process (the intermit- tent or continuous flow of fluids). These systems also relate to stages and eras, unit production in good part predating the Industrial Revolution, mass production being largely associated with it, and process production being largely a phenomenon of the twentieth century. Woodward found some marked relationships between these three systems of production and various of the design parameters. Specifically, in moving from unit to mass to process production:
- The span of control of the chief executives increased.
- The span of control of middle managers decreased.
- The ratio of managers to nonmanagers increased (from an average of 1 to 23, to 1 to 16, to 1 to 8); also, their qualifications rose (process organizations had more graduates, more managerial training, and more promotion from within).
- The ratio of clerical and administrative personnel to production personnel (indirect salaried to hourly paid) increased (from 1 to 1, to 4 to 1, to 9 to 1).
- The number of levels of management in the production depart- ment increased.
Moreover:
- The span of control of the first-line supervisors was highest in mass-production firms (about 48, compared with about 13 in pro- cess firms and 23 in unit-production firms).
- The mass-production firms had the smallest proportion of skilled workers.
- The mass-production firms were bureaucratic in structure, where- as the process- and unit-production firms tended to be organically structured.
But what distinguishes this study from the others is not these random observations but the way Woodward used them to paint an integrated picture of three distinctly different organizational structures associated with the three technical systems.
2. Unit production
The firms that manufactured individual units, prototypes, and large equip- ment in stages exhibited a number of characteristics in common. Most important, because their outputs were ad hoc or nonstandard, the unit producers’ operating work could likewise not be standardized or for-malized, and so their structures were organic. Any coordination that could not be handled by mutual adjustment among the operators themselves was resolved by direct supervision by the first-line managers. Being di- rectly responsible for production, the first-line managers worked closely with the operators, typically in small work groups. This resulted in a nar- row span of control at the first level of supervision. (The spans of control for the three different structures at three levels in the hierarchy are shown symbolically in Figure 6-2.) Woodward characterizes unit production as craft in nature, with the structure built around the skills of the workers in the operating core.
Figure 6-2. Spans of control at three levels in three techni- cal systems (based on the findings of Woodward, 1965)
These characteristics, in turn, meant little elaboration of the admin- istrative structure. With most of the coordination in the unit-production firms being ad hoc in nature, handled by mutual adjustment among the operators or direct supervision by the first-line managers, there was little need for an elaborate managerial hierarchy above them or a technostruc- ture beside them. Thus, of the three forms of production, the unit type had the smallest proportion of managers and, as can be seen in Figure 6-2, the widest span of control at the middle levels. At the strategic apex, however, the span of control tended to be narrow, a reflection perhaps of the ad hoc nature of the business. Not assured of a steady stream of orders, as in more routine production, the top managers had to spend more time with cus- tomers and so could not supervise as many people.
Mass production
If the structures of the unit-production firms were shaped by the nonstan- dard nature of their technical systems, those of the mass producers were shaped by the standard nature of theirs. Here mass standardized produc-tion led to formalized behavior, which led to all the characteristics of the classic bureaucracy. Operating work was routine, unskilled, and highly formalized. Such work required little direct supervision, resulting in wide spans of control for the first-line supervisors. The administration contained a fully developed technostructure to formalize the work. Woodward notes that the mass producers, unlike the other two, conformed to all the pat- terns of the traditional literature—clearly defined work duties, emphasis on written communication, unity of command, span of control at top levels often in the 5-7 range, a rigid separation of line and staff, and considerable action planning, long-range at the strategic apex (owing to the long prod- uct development cycles), short-range at lower levels (primarily to deal with sales fluctuations).
Moreover, Woodward found the structures of the mass-production firms to be the most segmented of the three and the most riddled with hostility and suspicion. She identifies three major points of conflict: (1) between the technical and social systems of the operating core, which gives rise to conflict that Woodward considers fundamentally irreconcilable, even in the well-run mass-production organization; (2) between the short- range focus of the lower-level managers and the long-range focus of the senior managers; (3) and between the line and staff groups in the admin- istrative structure, one with authority, the other with expertise.
Hunt (1970:171-72) refers to this second Woodward group as “perfor- mance” organizations, in contrast to the other two, which he calls “prob- lem-solving” organizations. In Hunt’s view, whereas the unit producers handled only exceptions and the process firms were concerned only with exceptions, the mass producers experienced fewer exceptions, these were of a less critical nature, and many of them could be handled by formal routines. These mass-production performance organizations spent their time fine-tuning their bureaucratic machines.
3. Process production
In firms built for the continuous production of fluid substances, Wood- ward found another structure again. What would cause these firms to be different from the mass producers? And why should Hunt describe them as problem solvers, concerned only with exceptions?
The answer seems to lie in a metamorphosis of structure when a technical system becomes so regulating that it approaches the state of automation. Mass production is often highly mechanized, but, if Wood- ward’s findings are a fair guide, seldom to the point of automation. The result is work that is highly regulated—simple, routine, and dull—requir- ing a large contingent of unskilled operators. And this, in turn, breeds an obsession with control in the administrative structure: supervisory, es- pecially technocratic, personnel are required to watch over and standardize the work of uninterested operators. With automation—which Woodward’s findings suggest to be more common in process production—comes a dra- matic reduction in the number of unskilled operators tied directly to the pace of production. Some giant oil refineries, for example, can be operated by six people, and even they only serve as monitors; the technical system runs itself.
With this change in the operating work force comes a dramatic change in structure: the operating core transcends a state of bureaucracy— in a sense, it becomes totally bureaucratic, totally standardized, but with- out the people—and the administration shifts its orientation completely. The rules, regulations, and standards are now built into machines, not workers. And machines never become alienated, no matter how demean- ing their work. So out goes the need for direct supervision and technocratic standardization and with it the obsession with control. And in comes a corps of technical specialists, to design the technical system and then main- tain it. In other words, automation brings a replacement in the operating core of unskilled workers directly tied to the technical system by skilled workers to maintain it, and in the middle levels of the structure a replace- ment of managers and technocratic staff who control the work of others by a support staff of professional designers who control their own work. And these changes dissolve many of the conflicts of the mass-production firms. Alienated operators no longer resist a control-obsessed management. Even at the strategic apex, “the company executives are increasingly concerned not with running today’s factory, but with designing tomorrow’s” (Simon, 1977:22-23). And staff need no longer battle line. This classical distinc- tion—between those who advise and those who choose—becomes irrele- vant when it is the control of machines that is at stake. Who gives orders to a machine, its staff designer or its line supervisor? Logically, decisions are taken by whoever has the specialized knowledge needed to make them, whether they be called line or staff.
With these points made, the Woodward findings about the process- production firms fall neatly into place, at least assuming that they are highly automated.4 She found that the process producers’ structures were generally organic in nature. Their operating cores consisted mostly of skilled, indirect workers, such as the service people who maintained the equipment. As in the unit-production firms, the first-level supervisory spans of control were narrow, again a reflection of the need for skilled operators to work in “small primary working groups.” This led to a “more intimate and informal” relationship between operator and supervisor than in the mass-production firms, “probably a contributing factor to better industrial relations” (p. 60).
Of Woodward’s three types, the process producers relied most on training and indoctrination, and had the highest ratios of administrators to operators, a reflection of the extensive use of support staff who designed the technical systems and also carried out functions such as research and development. They, too, tended to work in small groups—teams and task forces—hence the finding of narrow spans of control at middle levels as well. Woodward also found that the line/staff distinction was blurred in the process firms, it being “extremely difficult to distinguish between execu- tive and advisory responsibility” (p. 65). In some firms, the staff specialists were incorporated into the line structure; in others, “the line of command seemed to be disintegrating, executive responsibility being conferred on specialist staff” (p. 65). But it made little real difference; in any event, the line managers had training and knowledge similar to that of the staff spe- cialists, and the two in fact interchanged jobs regularly.
These firms also exhibited a sharp separation between product devel- opment and operations, resulting in a structure with two independent parts: an inner ring of operations with fixed facilities, short-range orienta- tion, and rigid control built into the machinery; and an outer ring of devel- opment—both product and process—with a very long-range orientation, loose control, and an emphasis on social relations. This two-part structure served to reduce conflict, first because it detached the technical and social systems from one another, unlike mass production, which put them into direct confrontation (here people could be free while machines were tightly controlled), and second, because it served to decouple the long- and short- range orientation. Another major source of conflict in the mass-production firms was reduced with the blurring of the line/staff distinction.
At the strategic apex of the process-production firms, Woodward found a tendency to use “management by committee” instead of by single decision makers. This was far less true of unit and mass producers. Yet she also found wide spans of control at the strategic apex, a finding that might be explained by the ability of the specialists lower down to make many key decisions, thereby freeing the top managers to supervise a large number of people. Perhaps the high-level committees served primarily to ensure coor- dination, by authorizing the choices made lower down.
To conclude, the dominant factor in the process-production firms Woodward studied seems to have been the automation of their technical systems. Automation appears to place an organization in a “postbureaucratic” state: the technical system is fully regulating, but of machines, not people, and the social system—largely outside the operating core—need not be controlled by rules and so can emerge as an organic structure, using mutual adjustment among the experts, encouraged by the liaison devices, to achieve coordination. Thus, the real difference between Woodward’s mass and process producers seems to be that although both sought to regulate their operating work, only the latter could automate it. In having to regulate people, the mass producers developed a control mentality that led to all kinds of conflict; in regulating machines, the pro- cess producers experienced less conflict.
With these findings in mind, we can now present three basic hypoth- eses about the relationships between structure and technical system.
Hypothesis 6: The more regulating the technical system, the more formalized the operating work and the more bureaucratic the structure of the operating core. As the technical system becomes more regulating—that is, broken down into simple, specialized tasks that remove discretion from those who have to use it—the operating work becomes more routine and predictable; as a result, it can more easily be specialized and formalized. Control becomes more impersonal, eventually mechanical, as staff analysts who design the work flow increasingly take power over it away from the unskilled workers who operate it and the managers who supervise them. We saw all these relationships clearly in Woodward’s mass-production firms. But what about those in process production? As Woodward de- scribed it, this technical system was almost completely regulating—that is, automated. Yet she characterized the structures of these firms as organic. But she meant the administrative structures, where the people were found. Their operating cores were, in a sense, almost perfectly bureaucratic; that is, in production (if not maintenance), their operating work was perfectly standardized; it just did not involve people.
Hypothesis 7: The more sophisticated (difficult to understand) the technical system, the more elaborate the nonoperating structure—specifi- cally, the larger and more professional the support staff, the greater the selective decentralization (to that staff), and the greater the use of liaison devices (to coordinate the work of that staff). If an organization is to use complex machinery, it must hire staff specialists who can understand that machinery, who can design, purchase, and modify it; it must give them considerable power to make decisions concerning that machinery; and they, in turn, must work in teams and task forces to make those decisions. Hence, we would expect organizations with sophisticated technical sys- tems to have a high proportion of support staff, to rely heavily on the liaison devices at middle levels, to favor small units there, and to decentral- ize selectively—that is, give the support staff power over the technical decisions. All these conclusions are suggested in the Woodward study; specifically, in the absence of an elaborate staff structure in the unit-pro- duction firms, generally with the least sophisticated technical systems, and in the presence of all these features in the process firms, generally with the most sophisticated technical systems.
Hypothesis 8: The automation of the operating core transforms a bureaucratic administrative structure into an organic one. We have al-ready discussed this hypothesis at some length in terms of Woodward’s process producers. Organizations dominated numerically by unskilled op- erators doing routine work are riddled with interpersonal conflicts. As Woodward notes, these stem largely from the inherent incompatibility of the social and technical systems: often, what is good for production is simply not good for the producer. As a result, mass-production firms de- velop an obsession with control—a belief that the workers must be con- stantly watched and pushed if they are to get their work done. Moreover, the control mentality spills over the operating core and affects all levels of the hierarchy, from the first level of supervision to the strategic apex. Control becomes the watchword of the organization. Top managers watch over middle managers, middle managers watch over operators and staff specialists, and staff specialists design systems to watch over everyone. Automation does not simply bring about more regulation of the activities of the operating core; as we saw, it eliminates the source of many of the social conflicts throughout the organization.5 Moreover, drawing on our last hypothesis, automated technical systems, typically being the most sophisticated, require the largest proportion of staff specialists. These peo- ple tend to communicate among each other informally and to rely for coordination on the liaison devices. And these, of course, are the most flexible of the design parameters. Thus, automation of the operating core breeds all kinds of changes in the administrative structure that drive it to the organic state.
This leads us to an interesting social implication: that one apparent solution to the problems of impersonal bureaucracy is not less regulation of operating tasks but more, to the point of automating them. Automation seems to humanize the traditional bureaucratic structure, something that democratization proves unable to do.
Source: Mintzberg Henry (1992), Structure in Fives: Designing Effective Organizations, Pearson; 1st edition.