Some problems of administrative theory

ORGANIZATION THEORY CAN BE APPROACHED in two ways. On the one hand, we can try, as in any science, to build up a factually correct description of the entities called organizations together with an explanation of their behavior, including the circumstances under which they behave effectively or ineffectively and how the effectiveness or ineffectiveness of particular organizational designs relates to the environments to which they must adapt. It is this “basic science” approach to organization theory that is taken in Chapter II.

On the other hand, we can think of organization theory as providing a guide to designing organizations—in the same way that architecture provides a guide to designing buildings, and engineering a guide to designing machines and structures. There is no conflict between basic science and engineering, but a notable difference in points of view. Science is concerned with establishing the laws that govern the behavior of systems of various kinds. Engineering is concerned with designing systems that will accomplish desired objectives. In this commentary, we will take a second look at organization theory from an engineering or design standpoint.


Consider the “proverbs” that occupied much of our attention in Chapter II. Classical theory asserted that an organization would be effective in so far as its design satisfied the “proverbs.” We have shown that these classical principles were mutually contradictory, hence did not provide a good base for a science unless we could determine, through research, under what circumstances and to what degree each should take precedence.

If we look at the same question, not as a matter of science but of engi- neering, it becomes less forbidding. From this new standpoint, the proverbs are not unbreakable laws but guidelines for design. For example: “When vnu are evaluating1 a scheme for specialization, consider to what extent activities aimed at the same goals are brought together, activities using the same processes, activities carried on at the same location, and so on.”

For designers to use these guidelines intelligently, they still need the scientific knowledge called for in Chapter II: knowledge of the circum- stances under which one or another guideline takes on special importance. The central difficulty in classical organization theory was its preoccupation with discovering categorical “principles” of organization, applicable unqualifiedly to all organizations at all times. The “principles of organization” still appear with great regularity in textbooks on organization, but they have gradually been relativized by a stream of criticism and empirical research that has shown that different organizational designs are needed for different functions in different environments.


As William Dill showed in an early study of this kind, a company that manufactures a wide range of diverse products for customers in several industries is bound to organize differently, if it wishes to survive and pros- per, from a company manufacturing a single line of products for a homo- geneous group of customers.21 Additional studies, which now provide a wealth of information about the adaptations of organization to environment, have been carried out by Joan Woodward, Tom Bums and George M. Stalker, Charles Perrow, James D. Thompson, Paul R. Lawrence and Jay W. Lorsch, and a number of others.9

Some of the research along these lines goes under the label of “con- tingency theory.” The central idea is that what constitutes effective orga- nization structure depends on goals and social and technical circumstances. This theme recurs throughout Administrative Behavior. For example, the commentary on Chapter XI addresses the relation of organizational form to environment and task. In the commentary, a case study of the Economic Cooperation Administration, the Federal agency organized m 1948 to administer the Marshall Plan of aid to Western European nations, provides a powerful illustration of how goals influence organization structure and vice versa. I will make a few preliminary com- ments on it here.

The ECA study emphasizes that designing an organization, like solving any other problem, begins with finding an appropriate way to represent the problem situation. Unless the designers come to the problem with a ready- made representation (i.e., the problem is of a kind that they have faced before), their initial concern must be to find such a representation, and only then can their attention shift to problem-solving.10 In the ECA, initial ambiguity of the agency’s goals and conflict among alternative goals led to the formation of competing representations. Only as these competitors were tested against the requirements of the agency’s task and a consensus was reached, did the organization take definite form. Stabilizing the decision process in an organization requires that most participants in the process.share a common picture of the organization and its goals.

A second example discussed in the commentary to Chapter XI is the organization of a business school, where one representation of the task comes from the sciences that underlie and inform business practice (e.g., economics, sociology, operations research, psychology, computer science); while a quite different representation comes from the “real” world of organizations and management to which the scientific knowledge is to be applied. What a business school, or any professional school, requires is a task representation that maintains a high degree of congruence between the pictures of the enterprise that are brought to it by faculty drawn from the world of science and faculty drawn from the world of practice.


Just as anatomy and physiology provide complementary approaches to the study of organisms, so structure and process provide complementary approaches to the study of organizations. Much of the research that relates organization to environment emphasizes the stable structural characteristics of organizations. In this volume we look more closely at the mechanisms of adaptation: how the decision-making process and the system of communications mediate between the organization and its environment. Two brief examples will illustrate how this viewpoint can be used to approach organization design in business.

Accounting Organization11

Some years ago an extensive study sought to determine how companies’ accounting systems should be organized in order to be of the greatest usefulness to executives in making their decisions and solving their problems. Answering this question required determining what important kinds of decisions were made by operating executives, how accounting data might be useful in making these decisions, and at what point in the decision- making process the data could most usefully be injected. By observing the actual decision-making process in detail, and in a number of corporations, specific data needs were identified at important organizational levels—the vice-presidential level, the level of the factory manager, and the level of the factory department head, for example—each posing distinct problems of communication for the accounting department.

Out of the analysis of data requirements at specific locations, a general pattern of accounting department organization was developed that would be effective in providing data for operating executives. For example, it was proposed to establish at the factory department level one or more accounting analysts, thoroughly conversant with operations, to help department heads interpret and trace costs through the monthly cost statements. At higher levels, on the other hand, it was proposed to create a small number of strategically placed groups of analysts largely occupied with special studies rather than periodic reports—analyzing the costs and savings associated with possible changes in operating methods and equipment.

Our present interest lies not so much in the study’s findings as in its implications for the technique of organizational design and reorganization.

  1. The foundation of the study was an examination of how decisions actually were made and where.
  2. The recommended organizational pattern for accounting was built around its task of informing and influencing these operating decisions.
  3. The recommendations for organizational change were to be imple- mented by bringing about changes in the patterns of who talks to whom, how often, about what—rather than by changes in organization chart.

Product Development

Industries that are based initially on a radically new technology typically go through several stages of product development and improvement. In the first stage, the principal source of product improvement is usually in the new technology itself and the sciences underlying it. Thus, when the computer industry was at this stage, industry leadership depended heavily on basic technical improvements in computer memories and circuits, improvements that stemmed, in turn, from advances in solid state physics and fundamental inquiry into the organization of computer hardware systems. At a later stage, product improvement became considerably more a matter of adaptation to end use—for example, providing appropriate software for customer applications.

An analysis of new-idea sources in these two stages would show that different types of research and development skills were needed, as well as different communications patterns between the engineering departments and their environments. In the long run, events forced the appropriate organizational changes on most companies (those that survived), but sys- tematic organizational analysis of the product development process could often have brought these changes about more promptly and profitably.

The main problems in organizing research and development lie in bringing together information from two distinct sources: from the scientific disciplines that underlie the basic technologies being used, and from the environments that define product requirements for the end use. But this brings us right back to the problem of organization mentioned in the previous section of this commentary: the organization of a professional school poses almost the same problem as does the organization of R&D: how to synthesize crucial information for decisions when it originates in different and remote sources.

As these examples suggest, the key method of analysis proposed in this book is to develop a careful and realistic picture of the decisions that are required for the organization’s activity, and of the flow of premises that contribute to these decisions. To do this, one needs a vocabulary and concepts that deal with organizational problems in a more fundamental way than does the homely wisdom that has passed for organizational analysis in the past.

Source: Simon Herbert A. (1997), Administrative Behavior, Free Press; Subsequent edition.

Leave a Reply

Your email address will not be published. Required fields are marked *