Printer Friendly

Organizational theory and organizational practice.

Scientific research, as you well know, is a process of exploration and discovery. But it is not a process of just looking around. It is disciplined, employing certain tools and methods including measures and observations that can be standardized and repeated so as to permit the test of hypothesis against reality. The methods of science, in other words, are designed to maximize objectivity.

There is both excitement and danger in this process of scientific exploration because one might discover what one is not looking for and in fact the scientist must assume that his theories are in some degree false and that the process of research is designed to find out just where the limitations of his theories are. Changes in scientific conceptions are therefore to be expected, and I would like this morning to trace some of the changes that have occurred during the past two decades in the thinking of some psychologists as a result of their explorations and discoveries in organizations.

Our problem is that of understanding the reactions of people in organizations; their conflicts, adjustments, motivations, satisfactions and dissatisfactions, and how these reactions affect the overall performance or effectiveness of the organization. As psychologists we approached this problem 15 or 20 years ago in a somewhat provincial way, assuming that the efficient organization is made up of efficient people. In other words, we focused attention on the individual and we looked, so to speak, inside the skin of the individual. For example, we proposed the hypothesis that a high level of productivity would be the result of a high level of job-satisfaction or morale on the part of the worker. This hypothesis was consistent with the human relations approach that was in vogue at the time, and we, along with other psychologists in the United States and Europe set out to test this relationship. To our disappointment we found that there was little support for the relationship. In most cases, the correlations found between morale and productivity were close to zero, and in some cases even, the relationships were negative.(2) The high producers were found in some cases to be those who had low morale -- which was, as you can imagine, somewhat shocking.

I would like to show you, just by way of illustration, one set of data which appeared ironic to us when they were first collected. This illustration is not offered as a typical result but it shows how research data may contradict a favorite hypothesis. Figure 1 presents the proportions of workers who participate in the recreation activities offered by a company. As the figure indicates, workers in high-producing groups are actually less likely to participate in the program than those in low-producing sections. These recreational activities, which were designed at some expense to boost the morale of employees, certainly did not have the desired effect on workers' productive efforts.

The problem with the hypothesis relating morale to productivity is that it is based on a rather naive psychology and a kind of Aristotelian logic: We said that job-satisfaction is a good thing and so is productivity; therefore productivity and job-satisfaction should go together. Unfortunately this hypothesis tells more about the values and prejudices of the researcher than it does about the phenomenon we were trying to understand. So much the worse for the hypothesis. The scientific method, as you see, was sufficiently objective in this case to overrule our own subjective feeling about how the world is, or should be structured.

So we began to explore elsewhere, including the worker's immediate social environment. We were interested in the worker's work group, including his peers and supervisor, and the effects these persons might have on the worker's adjustments and on his productive efforts. We collected data from a number of groups for purposes of analysis -- groups that were comparable in terms of their size, structure, technology and type of work performed, but which differed in their productive efficiency. For example, we asked workers and supervisors in high and low producing work groups how the supervisors in these groups behaved, and we found some rather interesting results. Supervisors in high producing work groups were reported to behave differently than supervisors in low producing work groups. In general -- and I'll have to summarize in the briefest possible way and at the risk of oversimplifying -- supervisors in high producing work groups were more considerate towards their subordinates; they took a personal interest in their men; they were less punitive in their attitude towards their subordinates; they supervised less closely rather than "breathing down the necks" of their subordinates; and they gave general instructions rather than very detailed and specific ones. Let me again for the sake of illustration, show you some data which came out of this part of the analysis. In Figure 2 we see the reports of subordinates about their supervisors, and we can see that 60% of the men in the high producing work groups say that their foremen are non-punitive and helpful while in the low producing work groups only 43% say that their foremen are non-punitive and helpful.

Now there are several problems with results of this kind. One is the problem of causation, namely what is cause and what is effect. For example, does the attitude of consideration on the part of the supervisor cause high production or is high production the cause of considerateness? An answer to this question requires experimental research of a type I shall discuss later. A second problem concerns application, namely how can these research findings be put to work improve organizational performance. How is it possible, in other words, to get more supervisors to behave in the superior ways suggested by this research?

The obvious answer to us at the time, again as psychologists, was training: Assemble the supervisors in a classroom, show them the research data, and illustrate the findings as principles of supervision. The results of these training programs, however, were disappointing. Supervisors, of course, were quick to learn right from wrong in the classroom situation, but it was another matter back on the shop floor. Many of the advocated principles of good supervision proved to be inconsistent with the circumstances in which the supervisor found himself. The supervisor is subject to many pressures, some of them from his own superiors and some from established organizational procedures and norms, that force him back into old habits. Figure 3 shows data from a study by Fleishman that illustrate this point. These data evaluate a supervisory training program that was designed to increase the considerateness of supervisors in their relations with subordinates. Part A of the figure shows the results of an opinion questionnaire that measures how the foreman thinks he should lead his work group. Part B shows the results of a questionnaire administered to the foreman's subordinates who describe the foreman's behavior on the job. The data refer to four comparable groups of foremen shown along the horizontal axis: those who had no training and those who had been out of training 2-10 months, 11-19 months, and 20-39 months respectively.

Fleishman points to the differences between the dashed and solid lines in these graphs. The former represent the consideration scores of supervisors who themselves work under leaders high in consideration. The solid lines apply to supervisors who are subject to a "leadership climate" low in consideration. These difference in leadership climate, not training, seem to be the more significant determinant of supervisory style.

A further problem with the research investigating the relationships between supervisory practices and the productive behavior of subordinates is the inconsistencies in results that were sometimes found from one study to another. In other words a number of conditions seemed to affect the extent to which the relationships occur. One such condition is the supervisor's power or influence in the organization hierarchy.(3) If an influential supervisor uses the appropriate techniques of supervision the effects on subordinates are likely to be favorable whereas if he uses inappropriate techniques, the effects are likely to be unfavorable. However, the uninfluential supervisor is less likely to have such effects; it makes relatively little difference what technique of supervision he uses.

We began as a result of such research to see the importance of conditions in the organization that effected the results of supervisory practices. It appeared that the attitudes and motivations of workers are affected in important ways not only by the supervisor but by the supervisor's superior and by key managerial personnel who may never be in direct contact with the worker. In many cases the supervisor is held accountable for the ill effects of decisions and actions which occur at levels above his own. Our research was leading us slowly but surely up the organizational hierarchy.

Let us take a look at some basic facts about this hierarchy. I want to show you some results from a study done during World War II in the Army of the United States. This study is well known and you may be familiar with some of its data. Figure 4 illustrates the interest that soldiers and officers in the Army had in their jobs. While these data were collected in the Army, they are typical of what is found in organizations: persons at higher ranks are generally more interested in or satisfied with their jobs than are persons at lower levels. Upper level persons are also more motivated, involved and personally identified with their work and they are more likely than those at lower levels to have a favorable attitude toward the organization as a whole.

Figure 5 provides another illustration, not from the Army but rather from industrial organizations. This figure presents data collected by Lyman Porter of the University of California from nearly 2000 executives. He asked these managerial personnel questions about the amount of self-esteem they derive from their jobs, their status, authority and other factors related to the job. Figure 5 shows the sense of deficiency which these executives feel with regard to their authority related needs, i. e., how much less authority they have than they feel they should have. As the figure shows, these persons perceive themselves as not having as much authority as they should have. Furthermore the felt deficiency is clearly greater for lower levels of management then for upper levels. Porter showed similar trends relative to other kinds of satisfactions as well: There is less of a sense of deficiency felt at the top of the organization than at lower managerial levels.

These data illustrate the impact on the individual member of the organization as a system. As a result of observations similar to these we began to focus on this system, on the larger context within which the individual organization member operates. It seemed reasonable to think that to change the behavior of a supervisor, it is necessary to create conditions in the organization as a whole, which make it appropriate for the supervisor to behave in the desired ways:

If it is desired to have supervisors treat their subordinates with consideration, to respect their men, it is not sufficient just to tell the supervisors to respect their men and to explain the rationale for this. Instead, it began to appear that the best way to get the supervisors to respect their men is to make the men respectable; that is, to change the organization in ways to give them some authority, responsibility, influence, control over significant aspects of their work life; to give them respectability. This requires change somewhat more extensive than that implied in the usual individually oriented human relations training program. In effect, one would have to "train" the entire organizational hierarchy from bottom to top as well as to change some of the working relationships including definitions of authority and responsibility at all levels in the organization.(4)

With this general idea in mind, Morse and Reimer performed an experiment in a large clerical organization.(5) The experiment took place in a department composed of 4 divisions each of which was doing the same work. In two of the divisions an attempt was made to delegate decision-making authority to lower levels in the hierarchy, so that employees with their supervisors could make decisions that affected their work lives. They could decide, for example, about work assignments work load vacation schedules, the length of recess, and about the right of workers to leave the department during the course of the day. The second pair of divisions underwent a contrary change in which control at upper levels was tightened and in which close supervision was increased. This experiment, which involved about 500 employees and lasted about 2 years, took place while the divisions were performing their regular duties, so it was not an artificial situation by any means.

The results were somewhat surprising. First of all, production went up in all divisions, but it went up somewhat more, about 25%, in the centralized divisions as compared to about 20% in the non-centralized divisions. This was distinctly contrary to expectation. However, other changes occurred that were more consistent with what we expected. Morale and job-satisfaction increased very distinctly in the decentralized divisions and decreased clearly in the centralized divisions. The motivational picture was also quite different in the respective divisions. In the decentralized groups, the sense of responsibility that the employees felt to get the work done on time increased whereas this sense of responsibility decreased in the centralized groups. Thus productivity increased in both experimental groups, but for different reasons. In the decentralized groups the clerks produced more because they wanted to: in the centralized divisions, they produced more because they had to.

A general implication of these findings was that there is more than one way to increase productivity. Productivity may be increased by tightening control from the top, but productivity may also be increased by placing greater control in the hands of persons at lower levels in the organization. Managers, however, are inclined towards the first, more traditional approach for several reasons. First of all the traditional approach is relatively simple and manifestly effective in terms of immediate productivity, as the data of this experiment show. But many managers favor the traditional approach for other reasons that we can state in the form of two assumptions. The first assumption we call the "all-or-none law of power": One either leads or is led, is strong or is weak, controls or is controlled. A second assumption, related to the first, is the "fixed-amount-of-power assumption", viz., that there is in any organization a fixed and limited amount of power so that increasing the power of one person or group implies decreasing that of others. We have reasons to question these assumptions and I would like to show you how we came to this questioning.

The two axes of the graph shows in Figure 6 represent two universal aspects of organization. The horizontal axis represents the hierarchical scale of an organization from the bottom to the top, and the vertical axis defines the amount of control exercised in the organization by the respective hierarchical echelons. While admittedly this graph is a simplification, it is nevertheless a useful device, as I hope to demonstrate. But first let us imagine a distribution of control like curve A, which is typical of business and industrial organizations, whose the greatest amount of power resides with the top echelons and little or none is at the bottom. Second one can imagine, theoretically at least, another kind of distribution of control -- something like curve D -- in which those at the lower end of the hierarchy as a group have more influence and power than those at higher levels. Other distributions also suggest themselves, at least theoretically (for example curves B and C) and these theoretical possibilities have led to several hypotheses which I shall illustrate in a moment. However, there are two aspects of this graph that I would like to highlight for your attention. One aspect is the distribution of control in an organization, illustrated by the slope of the curves. For example curve D has a positive slope as contrasted to curve A which is negatively sloped. The second aspect is the total amount of control in the organization which is represented by the average height of the curve, so that while the distributions are the same in curve B and C, the total amounts of control in the organizations represented by these curves are theoretically quite different.

This distinction between the distribution and the total amount of control in organizations suggested a resolution of some of the conflicting hypotheses with which we were confronted; the conflict between the value of increasing control by lower levels and the value of increasing control by upper levels. This resolution can be viewed as part of a dialectic: First we have the classical thesis or organization with control concentrated at the top, represented by curve A. Second we have the antithesis with control concentrated at the bottom, represented by curve B. Since these opposing positions have traditionally been stated in terms of the relative control of upper and lower levels, little attention has been given to the possibility of variations in the total amount of control exercised by the various echelons. The control graph, however, encourages consideration of this as the basis for a possible synthesis.

In Figure 7 we see the results of a study dealing with 31 departments of a large industrial organization. These departments have been divided into 3 groups two of which are shown in the figure. The dashed line curve shows data for the third of the departments highest in productivity whereas the solid line shows data for the third of the departments lowest in productivity. These graphs were obtained by asking employees in the respective departments to tell us how much influence various groups in their organization had over what goes on in their departments. The curves thus represent averages of the responses in the highest and lowest producing departments. We see that the distinction between the high producing and the low producing departments is not in terms of the relative control which upper levels exercise as compared to lower, but rather the total amount of control which these groups are reported to exercise in concert. We have tried this kind of investigation in a number of organizations and it is interesting to find in a good proportion of them substantiation for this kind of relationship.(6)

As you can see, these data are concerned as much with the organization as a system as they are concerned with the individual member within the organization. If we accept the data of Figure 7 as illustrative of an important factor of organization underlying the organization's performance then we are confronted again with the question of application: How can managers achieve the kind of organization system implicit in the control curves of the high producing departments of Figure 7? What techniques are available for changing the social character of an organization? Answers to these questions are now only being explored but I would like to mention some techniques that have been developed and are being developed that appear to be consistent with the style of effective organization, implicit in Figure 7. There is first of all a need to change the character of the total hierarchy and the way in which influence is exercised in the organization. In general, there is reason to think that some systems of participative management, contrary to the stereotype held by many, are means of increasing the influence not only of rank and file employees, but also that of managers.(7) An organization, as advocated by Likert and Mann, made up of groups as the basic building blocks, rather than individuals represents one such approach.(8) Each group, called an organizational family is composed of a supervisor and his immediate subordinates. These cohesive groups pervade the organization and provide important loci for the making of decisions which are pertinent to the work situations of the group members. As part of this system, supervisors are members of a second echelon of groups with their own supervisors. Supervisors, as members of two groups, serve to communicate between and help integrate the decisions of the groups. Such a system capitalizes on the high levels of mutual influence among members which is characteristic of cohesive groups.

The socio-technical approach advocated by social scientists at the Tavistock Institute is also consistent with this approach.(9) Socio-technical units are cohesive groups that have a sphere of authority delegated to them so that they can make decisions that apply to their work situations, -- decisions that they are in a position most appropriately to make. These groups are structured around meaningful technological units just as the technology is, designed in a manner consistent with the existence of the groups. In coal mining, for example, a system of cohesive teams, in which miners can support and help each other has proved superior to the system in which workers, because of technology, were relatively isolated. Socio-technical groups as well as organizational families are essentially a means for optimizing interaction and eliciting influence on the part of members so as to enhance the total amount of interaction and influence in the organization. Theoretically this should achieve a heightening of the curve of Figure 7.

Another approach to the heightening of the control curve has to do with training. We have had some disappointing experiences in human relations training, but there are developments today which may be more hopeful although they are still experimental. The training laboratory that employs the T group, and that attempts to work more intensively as well as extensively throughout the organization offers some hope. This training applies not only to supervisors who are taken out of the work situation and put in classrooms, but it is a training program that applies to people at all levels of the hierarchy who work together. The managerial grid approach of Blake and Mouton(10) and the attempts described by Argyris(11) to achieve interpersonal competence in an organization are consistent with the objective of enhancing influence throughout an organization through the use of laboratory training methods.

The method of "work simplification" developed by Allan Mogensen represents another technique which is consistent with the objective of enhancing influence throughout the organization.(12) "Work simplification" is especially interesting because it reconciles some of the important principles of human relations with those of scientific management. Essentially work simplification is a system in which workers and their supervisors or managers are formed into groups with the objective of creating new and better works methods. Group members investigate the production process and together they formulate ways of simplifying and improving that process; and this is a technological question as well as a question of human relations.

Workers' councils and joint management schemes may represent other approaches to the problem of enhancing influence in an organization although relatively little research evidence concerning the effects of these systems is now available. It is noteworthy therefore that studies investigating such approaches are taking place at two ends of the continent. In Yugoslavia, for example, there are studies of the effects of workers' councils on the distribution of control in firms, and in Norway at the other end of the continent research has been done on the system of workers being elected to boards of directors of some firms.(13)

Each of the above approaches to improving organizations is consistent in principle with the research we have described. Some of these approaches are only experimental; they are beginnings in the development and refinement of improved organizational forms. Doubtless, additional approaches will be developed as our conceptions of the human aspect of organization change, and as our understanding grows.

1 The main ideas, and the data presented in this talk are taken from the following sources, unless otherwise indicated: Katz, D., Maccoby, N., and Morse, Nancy. Productivity, supervision and morale in an office situation. Ann Arbor, Mich.: Institute for Social Research, 1950; Linert, R. New Patterns of Management, New York: McGraw-Hill Book Col, 1961; Tannenbaum, A. S. Social Psychology of the Work Organization, Belmont, Calif.: Wadsworth Publishing Co. (and London: Tavistock Publications) 1966: and Tannenbaum, A. S. and Seashore, S. E. "Individus et Organisations: evolution des conceptions et des modes d'analyse," Sociologie du Travail, 3, 1965, pp. 225-237.

2 Brayfield, H., and Crockett, W. H. "Employee attitudes and employee performance" Psychological Bulletin, 1955, 52, no. 5, pp. 396-424. Herzberg, F., Mausner, B., Peterson, R. O., and Capwell, Dora F. Job Attitudes: Review of research and opinion. Pittsburgh, Pa.: Psychological Service of Pittsburgh, 1957.

3 Pelz, D. C. "Influence: a key to effective leadership in the first-line supervisor", Personnel, Nov. 1952, pp. 3-11.

4 Tannenbaum and Seashore, op. cit.

5 Morse, Nancy C., and Reimer, Everett "The experimental change of a major organizational variable", Journal of Abnormal and Social Psychology, Vol. 52, No. 1, January, 1956, pp. 120 to 129.

6 See, e. g., Smith, Clagett, G., and Tannenbaum, A. S., "Organization control structure: A comparative analysis" Human Relations, Fall 1963, Vol. 16, pp. 299-346; Bowers, David G., "Organizational control in an insurance company" Sociometry, Vol. 27, No. 2, June 1964, pp. 230-244; and Bachmann, J., Smith, C., and Slesinger, J. "Control, performance and satisfaction: An analysis of structural and individual effects. Journal of Personality and Social Psychology, 1966 (in press).

7 March, J. G., and Simon, H. A. Organizations. New York: Wiley, 1958, p. 54. See also Burns and Stalker's description of "organic" as opposed to "mechanical" organization in The management of innovation. London: Tavistock Publications, Ltd., 1961.

8 Likert, R., op. cit.; and Mann, Floyd C. "Studying and creating change: A means to understanding social organization" Research in Human Relations. IRRA, Publication No. 17, 1957, pp. 146-167.

9 Emery, F. E., and Trist, E. L. Socio-technical systems. (Paper presented at the 6th Annual International Meeting of the Institute of Management Sciences. Paris: September, 1959.).

10 Blake, Robert, R., and Mouton, Jane, S. The Managerial Grid. Houston, Texas: Gulf Publishing Company, 1964.

11 Argyris, C. Interpersonal Competence and Organizational Effectiveness. Homewood, Ill.: Dorsey Press, 1962.

12 Goodwin, H. G. "Work simplification", Factory Management and Maintenance, 1958, pp. 72 to 106.

13 Zupanov, J. Grafikon Utjecaja Kao Analiticko Orude za Izucayanje Strukturalne Promjene Socijalne Organizacije Poduzeca. Zagreb, Yugoslavia: Ekonomski Institut Zagreb, 1964; Emery, F. E., and Thorsrud, E. Industrial Democracy, London: Tavistock Publications, Ltd., 1965.
COPYRIGHT 1992 Gabler Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Organization and Personnel
Author:Tannenbaum, Arnold S.
Publication:Management International Review
Date:Jan 1, 1992
Previous Article:System concepts: pervasiveness and potential.
Next Article:Some developments in the study of organizations.

Related Articles
Headquarters, host-culture and organizational influences on HRM policies and practices.
Ulrich Krystek, Wolfgang Redel, and Sebastian Reppegather: Grundzuge virtueller Organisationen -- Elemente und Erfolgsfaktoren, Chancen und Risiken,...
Human resource information systems: a review and model development.
Human resource configurations, intellectual capital, and organizational performance.
Human Resource Development Research Handbook: Linking Research and Practice.
Aptitude for Destruction: Organizational Learning in Terrorist Groups and Its Implications for Combating Terrorism, vol. 1.
Organizational behavior 4; from theory to practice.
Understanding action learning.
An innovative partnership of law enforcement and higher education.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters