Using data culled from over SO studies conducted over the past 10 years, CuTTtmrngs and Berger come up ivith ivhat, for the present, has to stand as the definitive answer to their question.

Organizatm Structure:
Mow T>oes Jt hfluenee Attitudes and Petfomame?
L. L. Cummings Chris J. Berger

he science of organizational analysis has now developed to the stage where numerous research studies—nearly 50 in the last decade alone—have examined the relation of organization structure to members' attitudes and behaviors. The history of organization and management theory clearly demonstrates a transition from the earlier prescriptive formulations of pioneer scholar-practitioners to a contemporary concern with description and even, in a few cases, with experimentation. A broad spectrum of organizational arrangements have been thought to affect

the satisfaction and performance of organizational participants. Among the organizational characteristics studied are; 1. The vertical level of the position held by an organizational member. 2. The nature of the authority (line versus staff) vested in the position. 3. The span of control (number of subordinates) of the position. 4. The size (number of members) of an organizational subunit—a department^ for example. 5. The size of the total organization.

Organizational Dynamics, Autumn 1916. © 1916, AMACOM, a division of American Management Associations. All rights reserved.

6. The number of levels of authority in the organization—whether its shape is relatively tall or flat. 7. The distribution of decisionmaking power across the levels of the organization—whether the organization is relatively centralized or decentralized. An approximate analogy would be to conceive of these characteristics as an organization's anatomy. They provide the structural foundation within which the organization functions. The question at hand is whether these structural features are related in any systematic manner to the attitudes and behaviors of people who are exposed to these features. It is true, of course, that prescriptive management theory has not only made the basic assumption that there are relationships but, in addition, has made two heroic assumptions. The first is that it is possible to be dogmatic about the ideal of some of the characteristics under all circumstances. Prescriptions exist, for example, that specify a span of control of seven, a work-group size of five, or a decentralized power allocation. The second is that these characteristics can be manipulated by managements at will in order to create highperforming, satisfying organizations. If these assumptions are to provide a valid basis for executive action, then it becomes crucial to know to what degree they are supported by systematic, cumulative research evidence. It is to this question that we now turn our attention. More than a decade ago, Lyman Porter and Edward Lawler published a review of the literature concerning the relationship of organizational structure characteristics to job attitudes and behaviors. They focused their review on the relationship between the seven organizational dimensions mentioned earlier and various employee reactions. The dependent variables

examined were job attitudes, broadly defined as an opinion concerning some aspect of one's job, and behavior relevant to the job: performance, turnover, absenteeism, and employee grievance rates. Most of the relationships that had been studied before 1964, the concluding date of the Porter and Lawler review, focused on attitudes rather than behavior. While the direction of the findings of the review was slightly mixed and far from conclusive, the general tenor was that higher-level positions, line