Organization Development In A Public Agency: The Strategy Of Organizational Devolution

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ORGANIZATION DEVELOPMENT IN A PUBLIC AGENCY The Strategy of Organizational Devolution by Larry Jack Kirkhart A Dissertation Presented to the FACULTY OF THE SCHOOL OF PUBLIC ADMINISTRATION UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements of the Degree DOCTOR OF PHILOSOPHY (Public Administration) February, 1973 INFORMATION TO USERS This dissertation was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted. The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction. 1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity. 2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will fjnd a good image of the page in the adjacent frame. 3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete. 4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced. University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 A Xerox Education Company I 73-14,419 KIRKHART, Larry Jack, 1938- ORGANIZATION DEVELOPMENT IN A PUBLIC AGENCY: TOE STRATEGY OF ORGANIZATIONAL DEVOLUTION. University of Southern California, Ph.D., 1973 Political Science, public administration University Microfilms. A XE R O X Company, Ann Arbor, Michigan THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. UNIVERSITY OF SO UTHERN CALIFORNIA THE GRADUATE SCHOOL U N IV E R S ITY PARK LOS ANOELES. C A L IF O R N IA 8 0 0 0 7 This dissertation, 'written by under the direction of h..f . ?„. Dissertation Com mittee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of requirements of the degree of D O C T O R O F P H IL O S O P H Y (/ Dtan DISSERTA IM IT T E E Chairman ... PLEASE NOTE: Some pages may have indistinct p rin t. Filmed as received. University Microfilms, A Xerox Education Company To Clare whose presence in my life made it possible to complete this study. ii TABLE OF CONTENTS LIST OF TABLES ....................... V LIST OF ILLUSTRATIONS Vi PREFACE vii CHAPTER I. INTRODUCTION 1 Theoretical and Practical Issues in Organizational Development and Design Models of Authority Structures and Processes of Change Focus of Study Summary II. ORGANIZATIONAL ENVIRONMENTS AND ORGANIZATIONAL Delineating Organizational/Environmental Relations Organizational Environments Organizational Structures The State Fund: A Case History of Organizational Evolution III. ORGANIZATIONAL DEVELOPMENT AND ORGANIZATIONAL The Concept of Organization Development Strategies of Organization Development Emergent/Situational Intervention Theory Applications of Emergent/Situational Theory OD Strategies: Retrospect and Prospect The Strategy of Organization Devolution DESIGN 29 DEVOLUTION 72 ill IV. A PARTICIPANT OBSERVER'S VIEW OF THE EXPERIMENT 136 Early Explorations in the Field Office Launching the Experiment A Major Unexpected Change Unanticipated Consequences of Crisis V. RESEARCH METHODS ........................... 162 Research Design Data-Gathering Strategy Statistical Analysis Summary VI. RESEARCH FINDINGS ........................... 193 Expectation of Leaders Task Characteristics Self at Work Summary VII. SUMMARY AND CONCLUSIONS..................... 232 APPENDIX.......................................... 24 5 BIBLIOGRAPHY ...................................... 275 iv LIST OF TABLES Table Page I. Total Populations of Respondents ............... 185 II. Respondent Populations after Equalization Procedures.....................................185 III. Variable 1: Rules Orientation......................197 IV. Variable 1: Rules Orientation......................198 V. Variable 2: Nurturance of Subordinates .......... 200 VI. Variable 2: Nurturance of Subordinates .......... 201 VII. Variable 3: Closeness of Supervision ............ 203 VIII. Variable 3: Closeness of Supervision ............ 204 IX. Variable 4: Universalism..........................206 X. Variable 4: Universalism..........................207 XI. Variable 5: Promotion-Achievement..................209 XII. Variable 5: Promotion-Achievement..................210 XIII. Variable 6: Participation in Decisions .......... 215 XIV. Variable 6: Participation in Decisions .......... 216 XV. Variable 7: Job Tension........................... 218 XVI. Variable 7: Job Tension........................... 219 XVII. Variable 8: Work-Related Change....................220 XVIII. Variable 8: Work-Related Change....................221 XIX. Variable 9: Self-Acceptance........................224 XX. Variable 9: Self-Acceptance........................225 XXI. Variable 10: Acceptance of Others.................226 XXII. Variable 10: Acceptance of Others.................227 v LIST OF ILLUSTRATIONS Paradigm Page I. Environmental Contexts ......................... 35 II. Organizational Designs ......................... 46 Figure 1. Uni-Hierarchical Competitive Authority .......... 21 2. Multi-Hierarchical Collaborative Authority .... 21 3. Consociated Model................................. 54 4. Organization-to-Environment Linkages............. 59 5. Organizational Power Patterns.....................125 6. Longitudinal Before and After Design with Aware and Unaware Control Groups ............. 165 7. Strategies for Minimizing Threats to Validity. . . 168 8. Threats to Internal Validity .................... 173 9. Summary of Research Findings.....................239 10. SCIF Organization Chart...........................241 vi PREFACE Like most long and detailed works, this research study could not have been written without the assistance and support of many, many people. The list is long as help and support were graciously offered by many col leagues and friends. In particular, I would like to express my gratitude to: Neely Gardner, a good friend who opened the door and provided unequivocal support; the members of SCIF, especially A1 Young and Frank Cancelliere, who insured maximum opportunity for the study; Bruce Storm, another good friend, who was patient and understanding; a friend named Bill Risconsin; Ann Strong (Newman) for performing her magic with the conputer in the early stages of data analysis; Jay Hansche for help in diving through the maze of possible statistical programs and landing with a good choice; Brian Donaher, who did a superb job of scanning and identifying resources; Guy Darst, who joined my "split infinitives" and tied my "dangling participles;" Lillian Christmas and Caroline Wadhams and her staff for their typing services; the Minnow- brookers in the hope that the New is not dead and the Old reincarnated; last and most inportantly, Clare, my wife, for straightening out the sea of problems created by a mind unalert to Kate Turabian and otherwise bringing order out of confusion. vii CHAPTER I INTRODUCTION Western civilization, particularly the highly industri alized sector, is afflicted by major paradoxes. Old and once legitimated political processes, essentially elitist, poly- archical systems, are more and more called into question and, at the same time, increasingly dominate. Bureaucratic organi zations, long the paradigm of formal organization, grow increasingly inefficient and ineffective. Yet, even though efficiency and effectiveness have conventionally justified uni-hierarchical organizations, and this justification no longer holds, few alternatives to bureaucracy have been attempted. People are faced simultaneously with widely proliferating choices, and with no choices which differ fundamentally from each other. We feel very little to be legitimate. What exists is all that is possible. All we can do is make incremental changes in the status quo. Public administration, because it is a profession, part of the structure of the social system, is one of these urgent concerns. Experiments in the design of public agencies are imperative. Interdependence among the various sectors of public administration and in the larger social system is great. 1 2 Consequently, it is difficult to find even a small corner in which to begin a real experiment and to continue it long enough to get results. For these reasons, the experiment in organi zational development and organizational design which is the basis of this dissertation is particularly significant. The study will develop a theoretical perspective to view the record of a unique experiment in organizational development. The experiment began within a limited part of a public organi zation and became an attempt to construct a viable, non-bureau crat ic public organization. In this introductory chapter I explore the general dimensions of the problem, identify its most complicated aspects, and give the reader an overview of what I discuss in detail in later chapters. Theoretical and Practical Issues in Organizational Development and Design One of the basic theoretical concerns which has arisen in a new and dramatic way during the last ten years is the relationship between an organization and its environment. Though this has been explored before, the relationship is now being examined in terms of its novel implications for the structure of organizations in a highly industrialized and turbu lent society. At both theoretical and empirical levels, a considerable amount of evidence suggests that this is indeed an issue we should take seriously.^ It has become increasingly clear that l the environment of an organization plays a very important role in shaping its potential for action. Organizational environ ments are known to affect: (1) the definition of organizational effectiveness, (2) the behavior which can occur within the organization and in relation to its environment, (3) the structure of authority in the organization, and (4) the character and quantity of human and technological resources which the organization can employ effectively. Whether genuine experimentation in organizational design is taking place depends largely on the extent to which the experimenters take into account the relationship between an organization and its environment. Moreover, at least two widely acclaimed authors insist that environmental turbulence and complexity requires great organizational flexibility and adaptability. Through theory and observations, they show that the adaptability which is increasingly demanded of organizations cannot be achieved through the two extant major forms of organization: Centralized bureaucracy or decentralized bureaucracy. The evolutionary soundness of uni-hierarchical, internally competitive organi zations may now be legitimately questioned. More than anyone else, Warren Bennis popularized the environment-organization issue in academic circles.3 Unfortu nately, his conceptualization embodied a major that now seems premature ancillary assumption: that environmental turbulence will inevitably produce the diametrical opposite of bureaucracy— non-hierarchical organization. Bennis' proposition, which departed radically from cultural experience in the United States,4 grew out of naivete about the complexity of transition from bureaucracy to other forms, as,well as from inadequate consideration of the circum stances that would most favor that transition. Undoubtedly the chief reason for this naivete was the lack of a well-developed theory of the relationship between organizations and their environments. It would, nevertheless, be a mistake to condemn Bennis' work because of this problem.^ This problem is grounded in the current status of organization theory. Consequently, a basic part of this study will be an attempt to develop a theory of the relationships between organi zational environments and organizational design.At a minimum, these theoretical components must be explored: 1. A conceptual framework which differentiates different types of organizational environments. 2. A theory of how different types of environments affect different types of organizations. 3. A testable framework for exploring the problem of what circumstances are, or are not, likely to encourage organizational change. 4. Alternate conceptualizations of formal organi zations which will facilitate identification and understanding. 5. Identification of which types of organizational/ environmental relationships are enabling and which ones are not. An enabling relationship is one which promotes positive change from the viewpoint of organization members and from the viewpoint of non-members affected by the change. 5 Relationships which prevent change or produce negative or disabling change must also be considered. Authority Structures in Organizations Organizational authority structures are centrally important to these large issues, and fundamental to experimen tation in organizational design. When Bennis popularized the topic, his position was that authority structures in organi zations do not and cannot exist independently of their cultural, social and technological context. Bennis' position is now very nearly axiomatic in contem porary theory. The literature of bureaucracy has been concerned with authority structures for decades. What we need now are 7 cogent formulations for non-bureaucratic organizations. Environmental turbulence may exacerbate the crisis of bureaucratic effectiveness. Turbulence may increase the allure of experimentation with new organizational forms. It is quite another matter to postulate that our deep-rooted cultural belief in the necessity of uni-hierarchical, competitive authority structures will evaporate under pressure of complexity and turbulence. A transformation of the bureaucratic system is unlikely to occur without considerable experimentation and widespread learning, a process that will require us to test the viability of different types of non-bureaucratic organization. One type of organization to supplant the uni-hierarchical' and competitive authority structure of bureaucracy would be a collaborative form of organization with several hierarchies.^ This form of organization and authority structure does not dictate to individuals the massive shift in values and the 9 increase in the basic security of their beings that the emergence of non-hierarchical forms would require. In an organization embedded in turbulences, movement of individuals toward an orientation that could support the emergence of multi-hierarchical collaborative authority structures involves, at a deep level, moving into a novel way of experiencing. It requires the emergence of a different feeling about acting within an organization. Since, in its most profound sense, feeling is being, it follows that feelings about one's role in an organization shape how a person acts as well as how others act toward the person. These transactions— individuals creating situations and acting in situations— form the basis of developing patterns of organization.^ Over time, the patterns become autonomous, legitimized institutional structures which persist despite the continuous trickle of personnel in and out of the 12 organization. Since feelings-in-situation are the basis of organi zational action, organizational change and development must somehow rest upon a change in feelings about involvement in the organization. Out of this change emerges the possibility of sustaining and developing new patterns of organizational action and structures. At present, we lack highly developed strategies that might guide this change. Process Theories of Organizational Change It is not accidental that over the past decade, the term "complex organization" has become commonplace. The phrase reflects, even somewhat understates, the enormous increase in our knowledge of the processes and multidimensionality of organizational behavior. Set against this large body of knowledge, the literature which realistically can be called contemporary theory of organizational change is scant. Only recently have we recognized the possibility of thinking about organizational change in more them one way.*'* But we have not clearly distinguished the several ways from each other and we have not developed the necessary terminology and concepts to make the distinction. Remarkably, we have not considered the organizational and environmental issues discussed earlier. Our theories are gross simplifications and never, as good theory should do, suggest how we might use them in the real world. Process Approaches to Organizational Change I noted above that feelings-in-situation are the basis of organizational action: this formulation led to a second proposition. Organizational change must rest upon a change in feelings about involvement in the organization. Taken together, these ideas raise the question of what strategies of change may affect feelings about involvement, thus encouraging or dis couraging organizational change. This conceptualization of the problem of organizational change is consistent with the issues I have discussed so far in this essay. The formulation is sufficiently general to 14 encompass both traditional and contemporary strategies of change. I will not, however, discuss or analyze here the traditional approach. Instead, I focus on contemporary change strategies which have developed in rough concurrence with the issues I have been examining. Two such strategies are apparent in the literature of organizational development and organizational change. Since these two approaches will be examined in detail below,^ I only outline them here. Grid Organization Development This theory and method of change, developed by Robert Blake and Jane Mouton,*® is designed to make members of an organization think about different types of managerial and interpersonal behavior. Blake and Mouton's managerial grid prescribes the most effective management style, known as 9/9 management, a behavior pattern that integrates concern for people and concern for production. These two concerns, for people and for production, are used to assess behavior within varied training designs devised to promote organizational change. Different training designs serve several purposes: management development, team develop ment and inter-group development. In all of Blake and Mouton's training, experiential exercises involve the person in assessing his own behavior. The critique made in relationship to the cognitive map provided by the authors is accomplished by completing paper and pencil tests and discussing the results with others in the organization. Careful analysis of the Blake and Mouton approach leads one to conclude that it is biased toward the development of uni-hierarchical, collaborative authority structures. It is humanistic in theory and technocratic in method of application. Emergent/Situational Intervention Theory A second approach which can be readily identified in the literature but is more difficult to describe may be called emergent/situational theory. This theory, short on the number of postulates embraced and long on openness and flexibility to situations, is most precisely defined in its essentials by Chris 17 Argyris' recent book, Intervention Theory and Method. This theory finds three conditions necessary to bring about organizational change: (1) valid information, (2) free choice, and (3) commitment. The strategy which follows from these postulates tries to create organizational situations in which these conditions exist. It is geared to accommodate whatever patterns develop as organizational members act within this set of conditions. Emergent/situational theory is used in the development of relations among persons, teams, and groups, as well as in organizational change. It is not biased against, or in favor of, any structure of authority except one that is congruent with the internal and external organizational situation. It is humanistic in theory and humanistic in application. Models of Authority Structures and Processes of Change Within the cluster of ideas termed emergent/situational theory, one author offers a singularly interesting formulation of two different type of authority structures.*8 Herbert Shepard describes each pattern in process terms and shows how they maintain themselves. The first process he describes is the way in which uni-hierarchical, competitive authority structures maintain themselves. The second describes the way non-bureau cratic, collaborative authority structures maintain themselves. Shepard discusses at length how each of these processes can change into the other. He begins by describing two different types of individual adaption. The first is called primary adaptation, which entails the following ideas: Survival of the self is impossible through others; the world is necessarily a competitive, zero-sum game in which relationships of trust with other human beings and mutually satisfying problem-solving activities are impossible; instrumental relationships are the only possibility. Freedom and responsibility are not identical. 11 This is called the primary mentality, an orientation to existence that produces certain principles of organization: 1. A pyramid of formal power and authority deals with coercion and compromise, which are the processes that provide order in an organization. 2. Each member of the organization agrees to let a higher level determine part of that member's behavior in exchange for certain resources needed for survival and well-being. 3. Emphasis within the organization (by members of the organization and the organization in relationship to its environment) stresses manipulating and gaining control over the environment to insure sur vival and well-being. 4. Members of the organization are induced to look upward for resources, downward to extract per formance, and sideways to compete more effectively with peers. 5. Intragroup solidarity and cohesiveness is main tained through intergroup competition. 6. All of the above is to be accomplished through a process of close, careful and vigilant adminis tration by a pyramidal authority. The author describes various negative processes these principles induce, such as distorted communication and lack of concern for the organization as a whole, producing ineffectiveness and inefficiency. Next, the author sets forth the secondary mentality and corresponding principles of organization. The secondary mentality assumes that individual well-being can be achieved through interpersonal relationships of a certain kind. Relationships that involve a high level of trust and mutual self-definition are both possible and desirable. Individual 12 survival and well-being need not conflict with the well-being and survival of others. The world is a non-zero sum game. From the viewpoint of the secondary mentality, individ ualism, collectivism, and totalitarianism are all consequences of a single view of the world. The primary mentality, defines the world as an inevitable, insoluble conflict between individual survival and the survival of others. To the secondary mentality, freedom and responsibility are identical. . . . Man lives most fully and freely when he lives responsibly. His growth to freedom is through expansion of self to include others, so that his concern for them is also his concern for himself.19 Shepard characterizes the organization principles which 20 flow from the secondary mentality as: 1. Commitment to a system of collaboration and con sensus to insure order in the organization. 2. Power structures, shaped to specific situations. 3. Agreed-upon goals which are defined through a process of open confrontation and dialogue. 4. Building structures which promote mutual confi dence and trust between members and groups within the organization. 5. A reward system tailored to a) commitment between organizational members and b) superordinate goals. 6. The form of the organization is not determined a priori; form follows function. These two sets of principles are essentially descriptive of the ways in which a bureaucratic authority structure and a non-bureaucratic authority structure maintain themselves. The author directly links these two structures and theorizes that 13 the secondary or non-bureaucratic structure can emerge from the primary structure. He describes a number of processes which must get under way, and events which must occur, in order for the non-bureaucratic structure to emerge. Shepard argues that the secondary pattern can and will emerge given the right conditions; his hypothesis connects the two patterns. At the same time, Shepard's view of how the process will unfold, or what kinds of conditions are likely to emerge in the transition, remains hazy. The hypothesis is nevertheless exciting, especially since his formulation is based on processes which support different types of authority structures. It is thus more susceptible to testing in a variety of circumstances. The problem is how to test the Shepard hypothesis under appropriate circumstances, as when the environment supports or facilitates change and the strategy for organizational change also encourages this secondary pattern of authority. Such a testing is the focus of this study. Focus of the Study This report presents empirical documentation of some effects, during the first two years, of a major, long-term organizational development program in a public agency. Specifi cally, I discuss the effects of the experiment upon the structure of authority in a field office. From the beginning, this program was based on novel ideas. There is a consequent need to identify the theoretical 14 and action components of the program. The strategy of change, organizational devolution, must also be related to or distin guished from the a priori and the emergent/situational 21 strategies discussed earlier. Organizational devolution springs from the idea that there is no a priori way to design a new, or develop an existing organization without continuous, thorough search and testing of the limits of the organizational situation. This idea was implemented through an experiment with a governmental organi zation. The organization was and is a relatively non-professionalized organization which operates in a highly complex environment. It sells workmen's compensation insurance to other state and local agencies and private firms. Its employees are under California civil service. Implementation of the organizational devolution strategy began with the level in the organization believed to be the primary determinant of the effectiveness of the system— the field office. The crux of the experiment is to establish circumstances in which the field office could define its need for services and support from the larger organization. Creation of these conditions was pursued by selecting a field office and granting it maximal autonomy and authority for a specified period of time. During this period, the field office was removed from the organizational hierarchy and could do anything that was within the scope of the prerogatives held by the General Manager of the overall organization. 15 Support and services required from parts of the organi zation external to the field office were to be obtained by contracting with the appropriate organizational units; the services were to be paid for out of the budget of the field office. Except for minimal information required by the State of California, the field office could discard any or all of the usual reporting systems. This essay will explore the social system of the experimental organization. Although, by definition, formal organization is a type of socio-technical system22 shaped in part by the nature and extent of the technology it uses, I shall assume that whatever change occurred during the experiment cannot be attributed to variance in the uses of technology by the organization. The functioning of the organization, both before and during the experiment, was affected mutually by the social system and the technological system. But because the technology utilized by the organization remained the same during our study, change or the lack of it was caused by the social system. The Research Problem and General Hypotheses Based on the theoretical issues discussed earlier, the novelty of the experimental conditions and their possible impli cations, the most compelling area of investigation should be clear. 16 The literature of organization development, and in the behavioral sciences generally, has documented many dimensions of organizational behavior: the effects of participatory management, the role of small groups, and, to a lesser degree, the consequences for organizational functioning of various types of intergroup relationships. The consequences of creating circumstances which enable systems to build authority patterns that suit their organi- 2 3 zational context have not been documented with similar precision. Because of the enabling strategy employed in this experiment, this study must document the realization of possi bilities. Trends and developments in the research data will be carefully examined. The research question is: Did the experiment increase the possibility of: the emergence of multiple, collaborative hierarchies within the organization? A generic definition of structures of authority in organi zations is central to the exploration of this question. The definition must remain meaningful even though relationships among its components change. Without this condition, we could not explore the Shepard hypothesis which supports the above research hypothesis. In short, the definition must be equally valid for several different types of authority structure. Considering these demands and drawing from the literature on the theory of organizations, I will use the 17 following variables to define organizational authority structures: 1. The type of leadership exercised within the formal role structure of the organization; 2. The way power, defined as the capacity of one person to influence another to act in a way desired by the first person, is expressed; 3. The legitimacy attributed to power holders in the situation; 4. The tendency of people to allow other people to define their reality; 5. The congruence of the basic organizational task with what the organizational actors define as worthwhile action; and 6. The constraints and requirements which flow from a larger social system and are compelling to a sub-system (personnel rules and regulations, promotional opportunities, technological imperatives, and administrative policies). The last variable will be the independent variable in this study; its manipulation was basic to the creation of the experimental situation. The other five items are the dependent variables. Each of these variables could be more narrowly defined. The problem is to develop more precise definitions that are both operational concepts and applicable to changing circum stances. For these reasons, the following definitions will be used: 1. Leadership Expectations.— Expectations about the role behavior of organizational personnel in leadership roles. These expectations are differentiated into the following areass^4 18 A. Rules Orientation.— "A normative expectation that prizes behavior which rigidly adheres to the formal rules of the organization." B. Nurturance of Subordinates.— An expectation that ''. . . a supervisor have at heart the best interests of those working for him and that he be solicitous in dealing with his subordinates.1 1 C. Closeness of Supervision.— The degree to which a supervisor delineatesand controls a subordinate's behavior. D. Universalism.— "A demand that the focal person disregard his primary relationships with his work associates, such as friendship or member ship in the same subgroup, in favor of loyalty to the social system as a whole, embodied here by 'the company.'" E. Promotion-Achievement.— A norm predicated on the belief that a person should be upwardly mobile and do anything that is likely to increase his mobility. 2. Power Utilization.— The degree to which it is perceived to be possible to influence various aspects of organizational situations. These aspects are 1) the work group, 2) the immediate supervisor, 3) the ease with which ideas about changing or improving a job can be communicated, 4) the way money is allocated to support the job, and 5) the upward influence of the supervisor.25 When influence along these lines is low, dominative power is utilized, when influence is high, collaborative power is utilized. This definition is keyed to the legitimacy of power. In other words, it encompasses two of the authority components— the way power is expressed and the legitimacy attributed to power holders. 3. Proclivity to have organizational reality defined by others. A. Job-Tension.— An anxiety state related to feeling unable to cope with the organizational task due to excessive load, lack of feedback about performance, conflict of roles.26 19 B. Self-Acceptance and Acceptance of Others^7 Self-Acceptance.— The self-accepting person: 1. Relies primarily upon internalized values and standards rather than on external pressure as a guide for his behavior. 2. Has faith in his capacity to cope with life. 3. Assumes responsibility for and accepts the consequences of his own behavior. 4. Accepts praise or criticism from others objectively. 5. Does not attempt to deny or distort any feelings, motives, limitations, abilities or favorable qualities which he sees in himself, but rather accepts all without se1f-condemnation. 6. Considers himself a person of worth on an equal plane with other persons. 7. Does not expect others to reject him whether he gives them any reason to reject him or not. 8. Does not regard himself as totally different from others, odd, or generally abnormal in his reactions. 9. Is not shy or self-conscious. Acceptance of Others.— The person who accepts others: 1. Does not reject, hate, or pass judgment against other persons when their behavior or standards seem to him to be contradictory to his own. 2. Does not attempt to dominate others. 3. Does not attempt to assume responsibility for others. 20 4. Does not deny the worth of others or their equality as persons with him. This does not imply equality in regard to specific achievements. He feels neither above nor below the people he meets. 5. Shows a desire to serve others. 6. Takes an active interest in others and shows a desire to create mutually satisfactory relations with them. 7. In attempting to advance his own welfare, is careful not to infringe on the rights of others. 4. Organizational Task and Value of Action.— The relationship between the nature of the organi zational task and its appropriateness, from the viewpoint of organizational members. This relationship is expressed through attitudes toward work-related change. Work-related change is defined as the degree of expressed willingness to change the definition of organizational tasks. It is negatively correlated with dominative power utilization.28 For simplicity of presentation, these variables will be regrouped into three general areas: Expectations of Leaders, Task Characteristics and Self at Work. Measurement of these indices in a uni-hierarchical authority structure and in a multi-hierarchical collaborative structure would produce quite different results. I show below what the pattern of the data would be. The general hypothesis is that, given the nature of the organizational development strategy and environment, the experi ment will produce a shift toward the second type of authority 29 structure. 21 Indices Measurement Dimension Measurement Scores Expectations Rules Orientation High of Leaders Nurturance of Subordinates Low Closeness of supervision High Universalism High Promotion-Achievement Low Task Charac Job-Tension High teristics Participation in Decisions Low Work-related Change Low Self at Work Self-Acceptance Low Acceptance of Others Low Fig. 1.— Uni-hierarchical competitive authority n Indices Measurement Dimension Measurement Scores Expectations Rules Orientation Low of Leaders Nurturance of Subordinates High Closeness of Supervision Low Universalism Low Promotion-Achievement High Task Charac Job-Tension Low teristics Participation in Decisions High Work-related Change High Self at Work Self-Acceptance High Acceptance of Others High Fig. 2.— Multi-hierarchical collaborative authority The specific hypotheses which will be tested are that the experiment will produce: 1. An increase in rules orientation. 2. An increase in nurturance of subordinates. 3. A decrease in closeness of supervision. 4. A decrease in universalism. 22 5. An increase in promotion-achievement expectations 6. An increase in participation in decisions. 7. A decrease : i-n job tension. 8. An increase in work-related change. 9. An increase in self-acceptance. 10. An increase in acceptance of others. Research Design In order to document carefully the effects of the experiment, I conducted a two-year longitudinal study. Such a study was required by the complexity and subtlety of the change process that might occur, by the lack in the literature of longitudinal studies of the long-term effects of organization development efforts, and by problems that arose during the experiment. I shall say more about each requirement in later chapters. I took data from seven organizations, including the experimental organization, four times during the two year period. This study will employ only the data collected from three of these organizations. One of these is the experimental organi zation; the second is an aware control; the third, selected from one of the remaining five, is an unaware control. It was necessary to take data from five organizations, in addition to the experimental and aware control organizations, to ensure that the unaware control was truly unaware. Because it was impossible to select at random the experimental and control groups, the research design is a matched-group design. Since the study was conducted in the field, where many potential sources of variance could not be controlled, the research situation must be termed quasi-experimental. Organization of the Balance of the Report In order to prepare the reader for later sections of the report, here is an outline of succeeding chapters: Chapter II.— An analytical perspective of some of the relationships between organizational environments and organi zational forms. The environment in which this study occurred is characterized. Chapter III.— An analysis of the two organizational development strategies currently practiced and the research literature upon them. These strategies are compared to the strategy of organizational devolution. Chapter IV.— Description of the experimental setting, including a chronology from the viewpoint of a participant- observer. Attention to events fund circumstances which seemed to shape the experiment. Chapter V.— An analysis of the research design, method ology, and data gathering techniques. Attention to the merits and limitations of the research tools and styles. Chapter VI.— Interpretation of the research data in relation to the hypotheses. Chapter VII.— A report of broader issues raised by the text and the significance of the research finding. 24 Recommendations for future research. Summary This chapter discusses the importance of experimentation in organizational design. The relationship between organi zational environments and organizational change is seen as especially significant, as are contemporary strategies of organizational change. The importance of organizational authority structures to the problem of organizational change is treated at some length. A process description of two different types of authority structures, bureaucratic and non-bureaucratic, follows; it is employed as the foundation of the study's general hypotheses. Operational definitions of the variables associated with organizational authority structures are established; how these variables will pattern themselves in bureaucratic and non-bureaucratic settings is outlined. Finally, I discuss the research design of the study and close with a chapter outline of the balance of the report. NOTES CHAPTER I 1. Warren Bennis, Changing Organizations (New York: McGraw-Hill Book Company, l96f); Paul Lawrence and Jay Lorsch, Organization and Environment (Cambridge, Mass.: Harvard University Press, 1967); E.F. Emery and E. L. Trist, "The Causal Texture of Organizational Environments,” Human Relations, XVII (February, 1965), 21-37; Robert Biller, "Adaption Capacity and Organizational Development," in Toward a New Public Administration, ed. by Frank Marini (Scranton, Pa.: Chandler Press, 1971). 2. Bennis, Changing Organizations; Herbert Shepard, "Innovation-Resisting ana Innovation-Producing Organizations," Journal of Business, XL (June, 1967), 267-81. 3. Bennis, Changing Organizations. 4. With the possible exception of temporary, micro-organizations such as encounter groups. Cf. Herbert Shepard, "Personal Growth Laboratories," Journal of Applied Behavioral Science, VI (November, 1970), 259-<>6. 5. Based on his own experience in carrying out his ideas he later published an article in which he pointed to the prematurity of his earlier formulation. Cf. Warren Bennis, "A Funny Thing Happened on the Way to the Future," American Psychologist, XXV (October, 1970). 6. Chap. ii. 7. In an earlier article, I attempted to work on this problem. Although I now am critical of my efforts to tie a non-bureaucratic form of organization, the consociated model, to the work of Max Weber, the basic model still seems theore tically cogent. Cf. Larry Kirkhart, "Public Administration and Selected Developments in the Social Sciences," in Marini, pp. 150-64. 8. Ibid. 9. R. D. Laing, The Divided Self (London: Tavistock Publications, 1960); Edward Tiryakian, "The Existential Self and the Person," in The Self and Social Interaction, ed. by ■ ■ 2 5 .......................... 26 Chad Gordon and Kenneth J. Gergen (New York: John Wiley & Sons, Inc., 1968), pp. 75-86. 10. Robert Janov, The Primal Scream (New York: G. P. Putnam's Sons, 1970); Frederick Peris, Gestalt Therapy Verbatim (Palo Alto, Calif.: Real People Press, 1970); Alexander Lowen, Pleasure (New York: Coward-McCann, Inc., 1970); Carl Rogers, On Becoming a Person (New York: Houqhton Mifflin Comoanv. 1959) . 11. Daniel Katz and Robert Kahn, The Social Psychology of Organizations (New York: John Wiley & Sons, Inc., l9(>6). These authors propose an event-structure theory which is con sistent with the transactional approach. 12. Peter Berger and Thomas Luckman, The Social Con struction of Reality (Garden City, N.Y.: Doubleday & Company, Inc., 19(56); Alfred Schutz, The Phenomenology of the Social World (Evanston, 111.: Northwestern University Press, 1967); Hans Peter Dreitzel, ed., Recent Sociology, Vol. II (New York: Macmillan Company, 1970); Karl Mannheim, Man and Society in an Age of Reconstruction (New York: Harcourt, Brace & Co., 1940); Kenneth Boulding, The Image (Ann Arbor, Mich.: University of Michigan Press, 1961). 13. Robert Chin, "The Utility of Systems Models and Development Models for Practitioners," in The Planning of Change, ed. by Warren Bennis, Kenneth Benne and Robert Chin (New York: Holt, Rinehart and Winston, 1961), pp. 201-14. 14. The essential features of traditional methods of change includes the following ideas. (1) Appropriate organizational change can be defined outside the context of the organization. (2) Uni-hierarchical, competitive (bureau cratic) organization is the only possible form of organization. (3) A derivative of the previous item is, there is one-best-way to organize and this way is consistent with the organizational theory literature. (4) The bureaucratic theory/literature is an appropriate characterization of the way to organize regardless of the nature of the context. (5) Organizational change is accomplished through the utilization of power-over relationships and is imposed from the top of the authority pyramid. (6) A singular pyramid of power and competitive processes to define upward mobility are essential to the main tenance of authority and order. (7) Organizational change is a technical, engineering problem; it is not, and cannot be justified as, a social and technical problem. 15. Chap. iii. 27 16. Robert Blake and Jane Mouton, The Managerial Grid (Houston, Texas: Gulf Publishing, 1964); Robert Blake and Jane Mouton, Building a Dynamic Organization Through Grid Organization Development (Reading, Mass.: Addison-Wesley Pub lishing Company, 1^69). 17. Chris Argyris, Intervention Theory and Method (Reading, Mass.: Addison-Wesley Publishing Company, 1976). 18. Herbert Shepard, "Changing Interpersonal and Group Processes," in Handbook of Organizations, ed. by James March (Chicago: Rand McNally & Company, 1965), pp. 1115-43. 19. Ibid., p. 1128. 20. Which are, incidentally, capable of producing compromise-coercion behaviors if the situation requires them but is otherwise committed to widening the possibilities of action. 21. Detailed comparison of these three strategies will be undertaken in chap. iii. 22. E. L. Trist, "On Socio-Technical Systems," in Bennis, et al., The Planning of Change, pp. 269-81. 23. This is a very different issue from documenting the effects of structural change imposed on the organization from the top down. This strategy is based on compliance, power-over relationships and deference to hierarchical authority. An extensive body of literature describes the consequences of this strategy. Within the top-down strategy, the nature of organi zational authority is a given and a constant. Within the enabling strategy, the possibility of developing new authority patterns exists if this is required by what organization members want to do. 24. Robert Kahn, et al., Organizational Stress (New York: John Wiley & Sons, Inc., 1964), pp. 153-57. 25. Victor Vroom, Some Personality Determinants of the Effects of Participation (Englewood Cliffs, N.J.: Prentice- Hall, Inc., 1966). 26. Kahn, et al., pp. 424-25. 27. Emmanuel Berger, "The Relation Between Expressed Acceptance of Self and Expressed Acceptance of Others," 28 Journal of Abnormal and Social Psychology, XLVII (April, 1952), 778-82. 28. Don Trumbo, "Individual and Group Correlates of Attitudes Toward Work-Related Change," Journal of Applied Psychology, XLV (August, 1961), 338-44. 2<f. Presentation of the rationale behind the hypotheses will be made in the chapter of research findings. This section is merely intended to introduce the general directions of the study. CHAPTER II ORGANIZATIONAL ENVIRONMENTS AND ORGANIZATIONAL DESIGN Change, environmental complexity, environmental turbu lence, technological complexity, technological imperatives, political change, social change, personal change, future shock, are all commonplace terms in both the academic literature and the popular press. Writers toss around cliches such as "we are living in the midst of accelerated environmental change" without hesitation. Organizations are uniformly assumed to be changing; if they are not, something is assumed to be wrong with the way they are managed. This unexamined bias can be seen very clearly in organi zation development. A body of literature has developed around the concept of the change agent. And the bias toward change has become so great that a prominent writer in the field has found it necessary to argue against emphasizing change for the sake of change. He points out that the power of the organization development orientation lies with its emphasis upon identifying and clarifying the circumstances of an organization here and now; it does not rest upon a priori assertion of the need for change.^ On the surface, this is a paradox. Consultants are warned against an unwarranted bias toward change precisely 30 when there is a growth in the number of organizations seeking out and hiring change agents (00 consultants) to help them become more effective in the process of change. This paradox suggests that the problem is more complex and subtle than has been acknowledged. The issue is, under what circumstances organizational change can be facilitated and under what circumstances can it not. There are at least two distinct ways to analyze this problem. The first is concerned with internal organizational processes; the second with relationships of the organization to its environment— the "ecology" of the organization. This essay employs the latter method of analysis. Delineating Organizational/Environmental Relations Given the prominence of systems theory, particularly open systems theory, in contemporary social science, it is curious that very little has been written on the relationship of an organization to its environment. Some writings talk about it in a general way but few offer any ideas about how to differentiate types of environments and what types of relation ships may exist between different types of environments and different types of organizations. Earlier in the text I pointed out that there are serious theoretical reasons for believing that these relationships have an important bearing upon the way an organizational system 2 behaves. One important by-product of the development of 31 theory in this area would be a more systematic and differen tiated way of organizing the existing literature on change, especially organizational change. Development of theory also would promote the possibility of a more sophisticated under standing of which contexts or situations lend themselves to the promotion of change and which do not; and this would, in turn, lead to a deeper awareness of specific processes in a given situation which have the effect of blocking or of facilitating change. I also argued earlier that a theory of the relation ships between organizational environment and organizational design would need, at the minimum, to contain the f o l l o w i n g :3 1. A conceptual framework which differentiates different types of organizational environ ments. 2. A theory of how different environments affect different types of organizations. 3. A testable framework for exploring the problem of under what circumstances organizations are, or are not, likely to change. 4. Alternative conceptualizations of formal organizations which will facilitate identification and understanding. 5. And, very importantly, identification of which types of organization-environment relationships are enabling and which ones are not. An enabling relationship is one which promotes positively regarded behavior from the viewpoint of organizational members and their social environment. Relationships which prevent change or produce negative or disabling change also need to be considered. 32 Organizational Environments Aside from the pioneering efforts of F. E. Emery and E. L. Trist, at the Tavistock Institute Human Resources Center, very little of the available literature lends itself to the theoretical concerns being discussed.4' 5' 6 Their purpose and the purpose of this essay is to simplify the problem of understanding, not to make the under standing more difficult. While at first this may not seem to be the case either in their article or in this essay, an impor tant point needs to be borne in mind; these essays are written against a backdrop of gross oversimplification; relative to this backdrop they are complex; relative to the inordinate complexity of the issue they are little more than primitive rules of thumb. In the material which follows I shall incorporate many of the concepts originally developed by Emery and Trist; at the same time, ray primary purpose is to extend the theory. I shall use many of the same basic concepts, develop new concepts and arguments, and, finally, present a paradigm which portrays four different types of environments. The next section of the essay will provide a similar kind of analysis of four different types of organizations and will include a paradigm that parallels the previous one. There will be a section which discusses the various relationships which theoretically could exist between the 33 various types of environments and types of organizations. Finally, I shall devote a section to the history and current status of the State Compensation Insurance Fund; I shall portray features of the environment which affect the organization and its design. I shall describe the theory of organizational environments and organizational design presented in the earlier part of the chapter. Four Types of Organizational Environments This analysis will proceed at the level of inter-organi- zational relationships; the boundaries of the analysis are more or less the boundaries of a nation-state; in other words, it will be a particular form of macro-social analysis. The per spective used is what this "world" looks like from the viewpoint of a particular organization. The conceptual scheme is based on four variables which are intended to represent generalized environmental conditions; each of these variables will be defined. These variables will be used to formulate four types of environments. The four environments are ideal-types; they are not meant to be a reflection of social reality; they are intended to be simplified ways of gaining a foothold on understanding social reality. Their construction rests on the following assumptions: (1) Reality is never completely unambiguous; therefore simplified schemes must be developed to ascertain an insight into a particular context; (2) A more profound knowledge of the world can be gained by segmenting parts of it and developing 34 unified constructs to represent these segmented parts; (3) Ideal-types must be constructed so that they can be utilized as a framework for comparing different empirical situations, b) the relationships posited in the type manifest causal adequacy, and c) the configuration or pattern of relationships is objectively possible.' The four variables are: Environmental Activity.— In some ways, this is the most straightforward of the variables; environmental activity means, simply, the degree to which energy is exchanged between organizations in a given society. If this energy exchange is low and does not have the effect of promoting change, this will be called a placid activity level. If, on the other hand, the energy exchanged between organizations is high and has the effect of creating new relationships, the possibility of new relationships and/or the abandonment of old relationships, or both, will be called a turbulent activity level. Shared Inter-Organizational Values.— This variable expresses the extent to which organizations in a particular environment have common values. A condition of high shared values can manifest itself two different ways. Organizations can share values in common but not recognize or acknowledge it; or, they can create, mutually share and acknowledge values; the latter condition offers the possibility of jointly addressing complex problems which effectively transcend the capability of a single organi zation. A condition of low shared values is said to exist when either the lack of inter-organizational contact or the conse quences of contact denies the possibility of shared value's. Interrelationship between Environmental Systems.— This phrasing is used to express the relationships between the analytical constructs which are commonly employed in contenpxary macro-analysis; these constructs are the political, social, economic and technological systems. Low interrelationshp is said to exist when the four systems overlap and are co-terminus with each other (functionally fused) or when all four are not interdependent. High interrelationship exists when all four systems are functionally distinct and interdependent. 35 Environmental Aggregation.— This is a measure of the possibility of organizational interaction within the environ ment. If the interaction potential is great, this will be called clustered aggregation; if it is not, random aggregation. Here is a paradigm which interrelates these four variables and yields four distinct types of environments. Following is an analysis of the main characteristics of each environment. Placid Turbulent Type II Type IV Clustered Industrial Post-Industrial Environmental Aggregation Type I Type III Random Early Industrial Late Industrial High Shared Inter- Organizaticnal Values Low Low High Interrelationship between Environmental Systems (political, social, economic and technological systems) Paradigm I.— Environmental contexts Before these environments are described, several important and limiting boundaries must be drawn. These boundaries are intended to clarify the context in which the different types of environments have theoretical utility and 36 those in which they do not. The framework is developed and intended for use in the context of Western Civilization and, in particular, the United States and, to a lesser extent, Canada. Qualification of usage such as this is necessary because the historical point at which industrialization began in a society affects both the particular nature of the environments which emerge with it and the possi bilities of new environments emerging. In short, this paradigm is not intended to be applicable to the "developing areas." These macro-social systems or nation-states face environmental conditions which Western societies also faced at the beginning of industrialization and they face important conditions which did not exist during earlier historical periods. Paramount among these are (1) the presence of a world-wide communications system, and (2) the influence of highly industrialized societies which sharply condition the process of industrialization The emergence of a world system and its penetration and influence upon the developing nations is something new and different from the forces which shaped earlier Western O civilization. This is not, however, to suggest that the international system does not, in turn, have an important effect upon the nature of environments in Western society. Clearly it does. The nature of this relationship is somewhat different. Inter dependence of the highly industrialized Western states and the international system is important; at the same time, the nature of this interdependence of states is quite different from what it was before the emergence of industrial countries because it is related to a situation ii. which the industrialized clearly have the dominant or greater amount of influence upon events— in most cases. The world system, as we shall see later, is an important component of Type IV or post-industrial environments and to a lesser extent Type III or late industrial environments. Type I: Early Industrial This particular form of environment is most charac teristic and discernible in the early phase of industrialization. In this setting, the location of formal organizations does not follow any particular pattern. Instead, they are more or less randomly scattered about the social terrain. Taken as a whole, inter-organizational interaction is the exception rather than the rule. The general character of this environment is, in comparison with the other types, placid and quiet. Political, social, economic and technological systems are differentiated in only minor ways. The dominant type of organization in this type of environment incorporates all of these systems and acts upon all of them regularly. Very generally, this form of environment was dominant from 1800 to 1900, with increasing reservations as one approaches the beginning of the twentieth century. In the literature of public administration, the celebrated Wilson article argues for a new form of 38 administration to cope with a new set of environmental forces-— even though he does not say this explicitly— and it is clear that his reference point is a Type I environment which was 9 rapidly being displaced by a different general environment. Contemporary existence of this form of environment is documented in a recent article which very clearly portrays the inter relationship of all of the various environmental systems (political, social, technological and economic) within the single dominant organization in the locale, the random presence and interaction of the various organizations in the context and the placid general character of the overall environment.*0 In general, organizations which compose this environment are located close to a particular type of resource; their purpose is the cultivation and development of this resource. This produces a unique organizational capacity and social identity. There tends to be very little interaction between organi zations and virtually no explicitly shared inter-organizational values. From the viewpoint of any given organization, the context is stable and predictable. Virtually any disturbances which threaten survival can be summarily dealt with and the status quo easily restored. Operations within organizations are also routine and repetitive and consequently are keyed to the development of long-term careers which in turn helps to maintain general environmental stability of housing, population, educat ion, etc. 39 Type II: Industrial Environments Type II environmental organization can be seen in a context in which the commitment to industrialization has already been made and the basic problem is completing the process in wider and wider sectors of the larger social system. A definite clustering of interacting organizations is present; from the perspective of any given organization, desirable and non-desirable forces in the general environment are clearly discernible. Each organization in the aggregate of organi zations which compose this type of environment has only random interactions with others and is attempting to reach a strategic location in the social system, that is, high status, which will ensure its long-term survival. For the first time, it is possible to identify two distinct different types of orientations which are pursued by individual organizations— strategy and tactics. In the Type I environment, there are no differentiation of these variables simply because the best way to deal with survival was to do the best one could in a particular setting. In the Type II environment, where organizational relationships are stronger and it is possible to identify positive and negative activity among organizational aggregates which affect a particular organization, knowledge of the environment becomes very important and provides the basis for choosing specific paths which would help reach the desired social location. 40 A primary vehicle for reaching this location is the development of a unique organizational identity which differentiates the organization from its potential predators. This is the essence of strategy in the Type II environment. Since the environment is complicated and interrelation ships between organizations fairly dynamic, it is difficult to know when to apply existing resources to a specific environ mental disturbance and when not to. This is the basic question of tactics--how, in the short run, to utilize resources most effectively. This is a troublesome and often crucial question because, typically, several organizations of a similar type exist and are pursuing similar goals. And, not only are long-term goals shared, there is also another basic and highly important set of shared values— the desirability and efficacy of competition and, thereby, survival of the fittest. The values are those of a zero sum game. All of this means that the specific tactics— short-run resource utilization— employed can be highly variable but are always keyed to the long-range goal of institutional survival. Interrelationships between environmental systems are distinct and sharply delineated. Political organizations are sharply distinguished from social and economic and technological systems. This holds despite the fact that organizations with these characteristics tend, increasingly, to be clustered in specific geographical locations— metropolitan or urban areas. 41 Nevertheless, each makes a sharp effort to distinguish its activities from those of other kinds of organizations. In many ways, this sharp institutional differentiation corresponds with the historical period of the politics-administration dichotomy and the tremendous emphasis upon the distinction between the public and private sectors of the society.** The specialization of function and separation of sectors produce a stabilizing effect upon the industrialization process and produce what can still be regarded as a placid level of environmental aggregation or organization. Type III; Late Industrial Environments The next order of complexity is represented by the Type III environment; political, social, economic and technological systems are much more closely interwoven. One of the signal characteristics of this environment is a continuously accelerating growth in the development of the technology and 12 its utilization. This growth is a consequence of larger and larger commitments to technological innovation as the key stone to organizational survival. Technology and a world-view associated with it, the technocratic perspective, produces a condition which can be described as unbridled technocracy.*^' **' *^ The political, social and economic systems are dominated and permeated by the imperatives of the utilization of technology. 42 Interrelationships between organizations are much more tightly woven than in either Type I or II environments;*6 at the same time, organizations deliberately randomize their relations with other organizations. Each organization attempts to keep its basic strategy for survival disguised from others and uses a variety of cooperative, competitive and obfuscating short-run tactics to enhance its maneuverability from situation 17 to situation. This has the effect of producing a large component of random signals between organizations; in the language of information theory, it is an environment charac- 1 f t terized by high noise levels. ° Because of the prevalence of this process of interaction, environmental aggregation can be accurately described as low. This is a rather curious phe nomenon, given the fact that it is no longer possible for any specific organization to exist wholly independent of others. Knowledge is widely shared; it is virtually impossible for any organization or set of organizations to maintain, 19 except for very short periods, information others do not have. Unlike Environments I and II, there are many organizations that are virtual replicas of each other. Each of these similar types of organizations is attempting to achieve the same societal position and, consequently, each tries to plant as many obstacles as possible in front of similar organizations. Environmental activity, taken as a whole, is intense but follows rather predictable and stable patterns. In 43 contrast to the total context, however, any given set or cluster of similar organizations is highly unstable and subject to change. Type IV: Post-Industrial Environments The identifying characteristic of this environment is that the entire environment field is changing. The instability and change present at the micro-level of organizational clusters in Environment III is present at the macro-level of Environment IV. Both macro- and micro-levels are highly dynamic and subject to disjointed and marked change. The dynamic properties arise not simply from the interaction of the component organizations, but also from the field itself. The "ground" is in motion. This movement is an unplanned consequence of continuous and widespread interaction of organizational sets and increased relations between the sets. Given the changing of the micro- and macro-environment surrounding a particular organization, there is an enormous growth in "organizationally relevant uncertainty." The consequences which flow from . . . actions lead off in ways that become increasingly unpre dictable; they do not necessarily fall off with distance, but may at any time be amplified beyond all expectation: similarly, lines of action that are strongly pursued may find themselves attenuated by emergent field forces.21 As was the case with Type III, technological growth and utilization occur at bewildering rates; in addition, there 44 is an upsurge of growth in political, social and economic systems. There is a distinct tendency for equalization to occur between each of these analytical systems as people in all sectors of the environment become more and more highly educated, professionalized and exposed to the workings of the whole system and its relationships with the world.22' 23 Behavior within sub-fields of the environment takes on a new quality; collaboration between organizations becomes the overriding imperative because problems are too complex to be 24 addressed from any other perspective than ecological. And, as effective functioning of the macro-field is increasingly dependent upon the functioning of sub-fields and vice versa, the demand for adaptation to secure survival shifts to the level of micro-fields; along with this shift there is an accompanying diminishment of the value associated with survival of individual organizations; social processes develop which support the ready transfer of personnel, technology and other resources within the field and enhance this process. This process is also supported by the increasing importance of collaboration between organizations. Turbulent fields, a name coined by Emery and Trist, is a very appropriate title for this environment.* Organizational Structures Since there is no one-to-one relationship between the demands of social reality and the development of the social science, it is not particularly surprising that when one of the most compelling aspects of society is its turbulence, the state of the art in organization theory (1) has not even adapted itself to the domain of inter-organizational relationships and (2) has hardly begun to touch the concept of turbulent fields. Since I myself am very much trapped by these shortcomings and dilemmas, this analysis will deliberately not address these two areas of concern but will focus on conceptualizing the response patterns a specific organization might generate to deal with each of the four types of environments. The responses described in the text are not the only possible patterns; they are simply one set of objective possibilities which are logical, given a specific type of environment. I argued earlier that there are some features of organization responses which seem to be a consequence of particular context, i.e., in Type II environments strategy is distinct from tactics; in Type IV environments the consequences of organizational behavior are hard to predict. One way to approach this problem is to again adopt a simplifying strategy. It should be borne in mind, that each of the organi zational designs in Paradigm II are nested in the environments described in Paradigm I earlier. Organizational Design II tends to be associated with Environment II, Design III with Environment III and so on. Since Environment I and Design I are the extreme limiting cases of "simple” organization, we shall begin with Design II, Environment II. 46 High Strength of Control over Total System Low Structural Adaptiveness Low High Design II Design IV Centralized Bureaucracy Non-Bureaucracy Long-Run Decision Design I Design III Perspective Pre-Bureaucracy Decentralized Bureaucracy Short-Run Centralized Decentralized Structure of Control System Paradigm II.— Organizational designs Design II Organization; Bureaucracy Because of the character of the environment, the primary demand which must be met for organizational survival i3 to achieve a particular location in societal space; the organi zation form tends to be centralized with a high level of total system control exerted from the top of a hierarchy; the hierarchy attempts to direct the activities of the organization in a way which will ensure that the identifiable and stable social location can be achieved. While there are occasional social disturbances in the environment, the organization can largely afford to pick and choose which disturbances or 47 variances will be recognized or which will be ignored. As the organization was originally designed to achieve a particular aid state, only minor structural modifications are expected to occur over time. Both the structure and the end-state are assumed to be stable; the problem is simply one of navigation from one point to another by moving a given organizational form to a desired societal location. Design III Organization; Decentralized Bureaucracy The environmental imperatives placed on Design III organizations demand that great attention be given to knowledge of the environment— if the organization is to survive over time. We saw earlier that characteristic patterns of behavior for a Design III organization to exhibit are (1) to send random signals to the environment in order to confuse similar organi zations about its real long-run strategy and (2) to launch major efforts to plant as many obstacles as possible before organizations pursuing the same long-run goal. This demands a structural flexibility that would be impossible to achieve under the Design II centralized organization. As the strength of the environmental challenge is great, there is a tendency for survival to be associated simply with the ability to move at will and for strategy to be subordinated to this imperative. Much attention is given to freeing the organization from its environment so that when it is necessary to "move" it will be possible; this pattern of concerns tends to increase the 48 importance of short-run decisions and diminish the attention given to long-run decisions. The rhetoric of long-run strategy remains while observable behavior is almost entirely oriented to short-run considerations. Design III organizations tend to be bureaucratic organi zations which have decentralized many of the control variables to low levels in the total organizational system but which have retained the power to exert overriding control of policy, personnel and economics. In some settings policy becomes the only overriding control mechanism evoked by top management, 2 7 and the other two variables are only very rarely utilized. In terms of overall effect, this provides for a flexible system which is capable of dealing with short-run decisions and effectiveness issues; long-run organizational strategy is in the hands of top management, well-removed from the imperatives of the field. This generates a planning-implementation vacuum; strong relationships between the long-run planning activities and field operations are highly improbable. The imperatives generated by the planning effort shape the organization in one direction and the cumulative effect of the expedient, short-run, field decisions is to set in motion consequences which also shape the actual strategy direction of the total organization— usually in ways that are either not recognized or poorly recognized by top management. Effective control is increasingly problematic. 49 Design IV Organization; Mon-BureaucratTc Earlier, in discussing the character of the environment called turbulent fields (Type IV) mention was made of the emergence of shared inter-organizational values; to take this point one step further, in Design IV organizations, a major shift occurs in the potency of internal organizational dynamics and environmental dynamics. All of the other organizational Designs, I, II, and III, attribute greater potency to internal functioning than to external dynamics. Design IV organizations, for the first time, place equal or greater reliance and importance on environmental dynamics; internal functioning is far from ignored, in fact it receives as much or more attention than in any of the other organizational forms. Self-initiated, extirpative behavior undertaken by organizations at appropriate times is socially prized as indicative of (1) a fully functioning, reality-centered organization and (2) an appreciation of and support for the survival imperatives of the overall society. Design IV organizational goals are intentionally defined in an open-ended manner; this places an acute emphasis on achievement of only those ends (1) which do not make an exaggerated claim to problem relevance over time and (2) which are designed to enable future decision-makers to have a maximum amount of flexibility. The aggregate of organizations and each organization within the aggregate rests on explicit time imperatives; in 50 other words, both the aggregate and individual organizations are committed to maintaining themselves only for specified time periods. There is no expectation of survival for the "foreseeable future"— rather, survival for a specifiable future. In short, this implies proactive emphasis on change and organizational design; virtually all existing change tech nologies and organizational designs rest on reactive premises; they are designed to help confront existing organizational problems and are not helpful in planning for currently non-existent but generally predictable dilemmas. Design IV organizations require a markedly different change strategy— one which is proactive— that incorporates a time limited plan for the future and is engaged today before a pressing need exists. In short, the present thinking about change is fitted to Organizational Design III, not Organi zational Design IV, to Environment III, not the turbulent fields of Environment IV. The dilemma in Organizational Design IV is not how to organize to cope with change but how to shape the future in the midst of organized, continuously pervasive change. Two conditions are crucial for the Design IV organi zation. The first is that societal support must be broader and stronger than heretofore; the values to which people, tech nology and knowledge are committed must be prized by social systems outside of the cluster of organically related organi zations addressing themselves to an ecological problem through 51 mutually shared values. The second condition is that new subsystems emerge within organizations and in the larger society; these social systems function as pre-changing centers designed to provide individuals with an opportunity to replenish their self-aware ness, to open themselves to a more fluid self-structure, in brief, to promote an optimal proclivity to learning. Organi zations such as these are not designed to be educational in the traditional sense but to provide personal growth opportunities, to provide non-threatening and minimally evaluative social surroundings in which one's becoming, to use Gordon Allport's term,can be explored effectively. The presence of these systems is crucial to the organizational effectiveness of Type IV designs. Otherwise, the temporary character of organized work settings would be very threatening and change-demoting. The possibility of both effective, experience-based organi zational learning and anticipatory problem-solving would be greatly reduced. If one knew, even though his role were to be temporary, the job was in accord with his own drives for becoming and that in the highly differentiated post-industrial society in which he lived he would have a maximum opportunity to become whatever it was possible for him to be, it is not likely that temporarily configured Design IV organizations would be seen negatively or as dysfunctionally stressful. 52 One of the standard features of a Design IV organization would be a high level of personnel turnover because as the organization moved closer to its goals the need for resources would be continually diminishing and dealt with accordingly. A large segment of the activities of the organization would be centered around finding and insuring a future job for the temporary residents in the system. The latter process incorporates one aspect of a drive for individual adaptability and systematic redundancy of functions within micro-fields. One has to have built-in redundancy or else settle for a system with a fixed repertoire of responses that are adaptive only to a finite, strictly identified set of environmental conditions. This is an important property of any system, as an arithmetical increase in redundancy tends to produce a logarithmic increase in reliability.29 This strategy, emphasizing individual adaptability, is a theoretically feasible way to ensure effective control of both the micro- and macro-environmental fields; central control systems would simply be unable to manage the complexity of 30 post-industrial environments. Complexity would quickly force a sustained elaboration of central control points and this would in turn logarithmically increase the probability that the components of the central control system would have inadequate information and be unable to achieve effective coordination. On the other hand, a commitment to individual adapta bility produces the possibility of problem-solving at the level 53 31 closest to the problem. Effectiveness Criteria: Design IV Organizations Organizational effectiveness for Design IV systems could be assessed along the following lines: 1. Speed with which viable relationships can be formed within and between organizations. 2. Timeliness with which organizations can build a set of viable, cogent shared values which consti tute the basic coordinative norms for a given set of interdependent organizations. 3. Accuracy of organizational goals in terms of how cogent they are to ecological problems and whether they are likely to generate dysfunctional consequences for contemporary or future decision- systerns and/or decision-makers. 4. Speed with which organizational goals can be reached without inducing undue societal disruption. 5. Accuracy of data shared with outside organizational aggregates and with society on achievement steps taken on the road to organizational goals. 6. Speed with which organizational sets and specific organizations in the sets can dissolve the system. 7. Care given to effective transfer of people, technology and knowledge to other locations in society and the speed with which these resources can be applied in other settings. 8. Allied with point 6, but sufficiently important to receive separate emphasis, is that the most highly effective organization will be one which maintains an accurate data flow from the environ ment which enables it to begin to readapt its resources, as the self-destruct process is underway, to be most easily related to new circumstances. In order to provide a clearer perspective on what a Design IV organization (that could meet these criteria) might 54 look like, the Consociated Model of organization is provided u i 32 below. The model is divided into two dimensions: Structural Characteristics and Social Consequences of Consociated Control. Structural Characteristics 1. Basic work unit is project team. a. financial autonomy 1) lump-sum budget b. situationally provided technology 1) tailor-made information and feedback system— computer based c. interdependent with other project teams 2. Multivalent authority structure. a. no permanent hierarchy b. leadership according to the situation c. diverse patterns of authority among various project teams 3. Total organization based on time imperatives. a. established to solve a particular problem within a specific time limit 1) subunits on shorter time parameters 4. Diverse subunit (projects) programs to deal with same basic problem. a. assumption of equifinality of means to goals 5. Social relationships characterized by high degree of independence and interdependence. a. direct, authentic interpersonal and iitergroup communications b. disintegration and redevelopment of project teams facilitates direct communications c. most permanent project team responsible for team building and basic encounter training 1) used to build trust, noncompetitiveness and interpersonal competence 2) team-building training prior to initiating project 3) basic encounter training used following disintegration of project team units Clientele served is represented in organization. a. cross section of clientele b. authority equal to professional members c. to obtain better information on service needs Organization is a place of temporary employment not a career. a. career associated with professional or vocational reference group outside organization Record keeping is computer based. a. performance goals of each project team shared with all others b. request for progress data automatically shared with others Professional role requires: a. technical skills b. skill in preventing emergence of more than minimal social stratification 1) promotes interdependence 2) promotes ambiguity useful in problem solving 3) maintenance of ambiguity builds greater trust between team members Social Consequences of Consociated Control Social diversity and independent personal styles are facilitated. Persons derive minimal alienation from organizational experience. Tolerance for conflict and ambiguity is increased. 56 4. Basis for joining formal organization is public values it supports. 5. Problem of structural rigidity is reduced through intentional elimination of organization. Fig. 3.— Consociated model This model is intended to be one example of a Design IV system which could perform effectively in post-industrial environ ments. A number of variations and no doubt a large number of other basic types could be just as possible. What the Formulation Is and Is Not Intended to Suggest- This exposition has explored four conceptually distinct environments, four organizational structures or designs and has suggested that specific designs are consistent with specific environments; no assumption has been made that these relationships would necessarily be descriptive of social reality. This theory is a simplification of reality (not a mirror) to be used as a tool for assessing organization-environ- ment problems and to provide some predictions about what kinds of organization development programs are most likely to be successful and what kinds are not. Even this proposition is an oversimplification for, as noted in the text, in complex, highly industrialized societies like the United States and Canada, aspects of all four environments and all four organi zation designs exist simultaneously. 57 In the context in which two or more of these environ ments co-exist, the problem of organizational design is inordinately more complicated. And, it is also likely that the classical bureaucratic and depersonalized form of organization, while retaining its basic form, will entail more internal turbulence and concern for its human resources than was the case historically. J However, it is well beyond the scope and purposes of this essay to pursue a more detailed analysis than has been presented so far. Based on the theory, we would predict that if we launched an organization development program to change a Design II organization to a Design IV organization when that organization was nested in a Type II environment, we would fail. Or, if an organization was helped to reorder its structure from Design II to Design III and the organization is almost entirely confronted with Type IV environment, the strategy of development, even if the change was brought about, would not increase organizational effectiveness— in fact, in the short run, it may induce greater potentiality for social obsolescence and ineffectiveness; Design III organizations do not have a highly adaptable control system; primary controls rest within one location in the organization and a large segment of control activities lies outside this primary center— all of which implies that in the highly dynamic turbu lent fields environment (IV), the possibility of effective adaptation to external demands is not great. Secondly, no assumption is made that either the environ mental types or the organizational designs rests on a continuum. Each of the types is representative of a specific set of relationships which are objectively possible; they do not violate existing knowledge of human action in situation. Thus, contrary to the dictates generated by "thinking along a continuum," the formulation used in this essay does not preclude the possibility that a specific organization could change from Design II to Design IV without moving through a stage of development comparable to Design III. There is no necessity of thinking about developmental stages as one is led to do when patterns of behavior or social systems are viewed on a continuum. Without these constraints, the problem of the relation ship between organizational environment and organizational design can be explored in a manner that leads to many possi bilities. Using only the four environmental types and the four organizational forms described earlier, we can examine a large number of relationships affecting organizational performance. It is possible to distinguish between environmental characteristics and the character of environmental imperatives placed on an organization; simply because a specific type of environment surrounds an organization, there is no reason to believe the demands of this environment necessarily will be accurately understood and acted upon by the organization. It is also possible, even if an organizational system does possess 59 accurate knowledge of the environment it faces, the responses the organization makes to the environment or the linkages it builds with the environment may not be congruent with the environment. Thus instead of thinking only about environmental types and organizational designs, we cam add two additional considerations: environmental imperatives placed on the organization and organization-to-environment linkages. Environmental Type Environmental Imperatives Placed on Organization Organization to Environ ment Linkages Organizational Design I I I I II II II II III III III III IV IV IV IV Fig. 4.— Organization-to-environment linkages We would predict that organizational development efforts have greater potential success in settings where there is an incongruence in any or all of these four relationships. This provides a clearer focus on the reason why organizational adaptability has become a major issue. As we said earlier, organizations in the United States and Canada face all four types of environments simultaneously; consequently, the stresses placed on any organizational design are enormous. 60 Summary Considerable effort has been made to develop a theory of organizational environments and organizational design. Four types of environments were described and organizational designs which were theoretically consistent with them were discussed. Particular attention was given to post-industrial environments and the demands on performance organizations face in this context. Criteria of organizational effectiveness for Design IV organizations were developed and a model of non-bureaucratic, Design IV organization was presented. This model was, in part, developed on the basis of experiences the author had with the organization that is the focus of the overall report— the California State Compensation Insurance Fund. The next section will present a chronology of the Fund and the evolution of environmental demands it has faced and the forms of organization which were developed to cope with these demands. The State Fund: A Case History of Organizational Evolution The California State Compensation Insurance Fund (SCIF), unlike most state organizations, does not directly depend on the budgetary decisions of the Governor and the Legislature for its money. In 1914 the legislature authorized the establish* ment of a revolving fund of $100,000 to found the organization; 61 this money was soon repaid and since that time SCIF has been financially independent. In these early days, SCIF had an important role in setting standards for the insurance industry in the field of workmen's compensation. It was known then— and has maintained this reputation— as a progressive, efficient government organization.^ Its operations have always involved both the public and the private sectors. Under California law, all public agencies that are not self-insured must purchase insurance from SCIF. Besides holding this quasi-monopolistic market, the organization always has written insurance in the open-market in direct competition with private companies. From the beginning the organization was a highly centralized bureaucratic structure (Design II); this structure was maintained without major modification until approximately 1960. At that time, top management decided, on the basis of research conducted by Stanford University, that it should decentralize many of its operations. During the next ten years a major effort was made to provide greater responsibility for making decisions to lower levels of the organization, i.e., Design III organization. In recent years severe stress has been exerted on SCIF by both the public and private markets. Many public agencies came to believe that neither the quality of service they received from SCIF nor the prices they paid were as good as they could get by insuring themselves or by buying insurance 62 in the private market. Such dissatisfaction prompted interest in or movement toward self insurance. This revolt of public agencies has been openly supported by the private insurance carriers; historically, private insurance companies have lobbied and politicked at the State level to have the State Fund abolished. With the advent of many governmental organizations which can choose to insure themselves instead of insuring with the Fund, new political forces now exert greater, more complex pressure upon the State Fund organization. A second major environmental imperative was created by organizational amalgamation. More and more private carriers were becoming national organizations capable of serving branch offices of corporations across the United States. The Fund cannot do this; State regulations limit all SCIF activities, including sales to the State of California. This competitive problem was aggravated by the fact that SCIF's client organizations too were often subsumed by amalgamated national organizations. Unless clients were offered superb service, highly efficient claims adjustment and low premiums in a particular locale, their national headquarters would buy workmen's compensation insurance from a single carrier that could serve all the organization's sub-systems nationwide. In the process, the client corporation could usually expect a significant overall 63 premium decrease because of economies of scale. As a generous proportion of the SCIF's clientele locate major headquarters outside the State of California, maintaining the accounts of local branches of national organizations has become increasingly difficult for SCIF. Issues Within the Organization A second set of imperatives emerged from within the organization. As of July, 1967, the organization was divided into two divisions--Northern and Southern. Within each division were approximately twelve District Offices. Each District Office is a nearly self-contained unit whose personnel are arranged in the following categories— sales, safety, auditing, clerical and, in many instances, claims. Generally each office employs about 12 salesmen, 6 persons responsible for improving the safety standards of the larger accounts, 6 auditors who periodically review the books of the clientele, a staff of about 5 clerks and clerk-typists who serve the others and, in the case of the dispersed offices, generally about 12 people who manage all insurance claims made to the organization, except when litigation is required. Each organi zation has a legal services department that is independent of the line of authority within the District Office. District Offices without claims units are known as Non-Dispersed Offices. During recent years, SCIF has moved to decentralize the organization assigning the claims function 64 to the field offices. By the summer of 1969, all of the field offices incorporated a claims unit. One of the most pointed characteristics of SCIF is its belief in generalist training. This can be most clearly observed at the level of the District Offices, where beneath the level of District Manager each stratum is constantly rotated so that each person acquires some expertise in each of the organization's four primary functional activities (excluding the clerical unit). In addition, most supervisors in the District Offices, almost all District Managers, and managers more advanced in the hierarchy have at some time participated in human relations laboratories. Currently, a program based on team development or "family" labs is progressing in several District Offices. The top administrative position in the District Offices was that of District Manager.(DM). Below him are sales supervisors, safety supervisors, auditing supervisors, clerical supervisors, and claims supervisors, all stipulated in SCIF's job categories. Above the level of the District Manager was a Regional Manager responsible for approximately half (generally about six offices) of the District Offices in his division. The two Regional Managers reported to one of the two Division Chiefs, North or South. At the level of Division Chief there were, in addition to the two who oversaw field operations, several other Division Chiefs responsible for various staff services— 65 Underwriting, Insurance Services, etc. Division Chiefs reported to the Assistant General Manager, who reported to the General Manager. The chief internal problem of the organization was that District Offices faced situations that demanded increasingly high levels of adaptability. Although the two Divisions differed in their degree of awareness of general operating policies as developed by top management, few District Offices found that these policies (to the extent that they were known) provided helpful guidance in their day-to-day operations. Many field personnel, particularly those in the Southern Division were uncomfortable with the prevailing level of ambiguity regarding appropriate operating procedure as defined by the organizational hierarchy. In an ambiguous situation field people could not know how the next level of the organization might react to decisions they had made; nor would field offices know how long to wait for higher management's advice after asking it. As both speed and cogency of response were acutely felt needs in District Offices, it is not surprising that many DMs were highly dissatisfied with the response of the organi zational hierarchy above them. As the field offices demand for greater responsiveness grew urgent, the top of the organization began to ask for better feedback about processes, problems and prospects within the organization. One of the problems faced by the General Manager was that he received little accurate, high quality 66 information about the operations of the organization; by the time it reached the General Manager, most information had been filtered through so many levels that it was both dated and obscure. A demand for structural change was therefore growing at both ends of the organizational pyramid. At this point, Neely Gardner was hired as a process consultant to help gather information about the organization's functioning. Professor Gardner had worked as an organization development consultant with top SCIF management personnel for a number of years. His activities included team-building, management training, and process consultation during many top management meetings. Nevertheless, the research we are considering and Gardner's role in helping management to process the information, was crucial to subsequent events. Mr. Gardner interviewed a large number of DMs, the Division Chiefs, and many of the staff personnel in SCIF's central office in San Francisco. On the basis of the data gathered and the pressures toward change registered inside the organization, top management decided to undertake an experiment directed toward giving field offices optimal maneuverability and decision-making power. Early in the summer of 1967, it was decided that one District Office would be chosen for an experiment in organi zational autonomy; a second office would continue to operate in the traditional manner and be used as a control group to 67 measure the performance of the experimental office. The second office would be told it would be used as a reference point. A third District Office would be used as an unaware control group, operating under the old procedure without being told it was involved in the research. All of the offices selected for the experiment were similar— Non-Dispersed Offices in comparable environments. Data gathered from this experiment would be used to develop a more effective organizational design— one suited to the turbulent environment of the organization. NOTES CHAPTER II 1. Chris Argyris, Intervention Theory and Method (Reading, Mass.: Addison-Wesley Publishing Company, 1970). 2. Cf. Chapter i. 3. Chapter i, p. 4. 4. A second and somewhat less cogent article is E. L. Trist, "The Relation of Welfare and Developments in the Transition to Post-Industrialism," (Paper presented at the Canadian Center for Community Studies, Ottawa Seminar, November, 1967), 62 pp. 5. One additional piece of literature contains material which is directly relevant; this article is based on the Emery and Trist schema and contains a superb review of the literature in organization theory which, in any way, addresses this area. Unfortunately, the article is premised on the thoroughly discredited idea of the necessity of linear evolution in social systems. Cf. Shirley Terreberry, "The Evolution of Organizational Environments," Administrative Science Quarterly, XII (March, 1968), 590-613. 6. With one major exception, the basic types of environments identified by Emery and Trist are not unknown, if the entire body of literature in the social sciences is taken into consideration. Three of their environmental types have features which have been extensively discussed in economic theory, i.e., the market, monopolistic and oligopolistic models. It is, nevertheless, rare to find an analysis which even attempts to draw from such a broad base. Their chief contributions are: (1) the conceptualization of a new type of environment, (2) the development of an analyt ical method of examining organization-environment relations, and (3) a novel way of looking at environments. 7. Larry Kirkhart, "Public Administration and Selected Developments in the Social Sciences," in Toward a New Public Administration, ed. by Frank Marini (Scranton, Pa.: Chandler Press, 1971). 68 .............. 69 8. R. Buckminster Fuller, Operating Manual for Spaceship Earth (Carbondale, 111.: Southern Illinois University Press, 1969); James N. Roseneau, "Pre-Theories and Theories of Foreign Policy," in Approaches to Comparative and International Politics, ed. by R. Barry Farrell (Evanston, 111. : Northwestern University Press, 1966), pp. 27-92. 9. Woodrow Wilson, "The Study of Administration," Political Science Quarterly, II (June, 1887), 197-222. 10. Although clearly not a scientific study, this article portrays important aspects of a Type I Environment as it exists in a contemporary setting. Peter Schuck and Harrison Wellford, "Democracy and the Good Life in a Company Town," Harper's Magazine.CCXLIV (May, 1972), 56-66. 11. Cf. Dwight Waldo, The Administrative State (New York: Roland Press, 1947) . 12. Jacques Ellul, The Technological Society (New York: Alfred A. Knopf, Inc., 1964). 13. Theodore Roszak, The Making of a Counter Culture (Garden City, N.Y.: Doubleday & Company, Inc., 1969), pp. 1-41; 205-238. 14. Anton C. Zijderveld, The Abstract Society (Garden City, N.Y.: Doubleday & Company, Inc., 197 0)1 15. Trent Schroyer, Toward a Critical Theory for Advanced Industrial Society," in Recent Sociology, Vol. II, ed. by Hans Peter Dreitzel (New York: Macmillan Company, 1970), pp. 210-234. 16. Robert Presthus, The Organizational Society (New York: Random House, Inc.7 1965), pp. S$-$2. 17. Although its title seems erroneous, the following book contains an assessment that is similar: Alain Touraine, The Post-Industrial Society (New York: Random House, Inc., 1971). 18. Philip M. Hauser, "The Chaotic Society," American Sociological Review, XXXIV (September, 1969), 1-19. 19. John Kenneth Galbraith, The New Industrial State (Boston: Houghton Mifflin Company, 1967) . 20. F. E. Emery and E. L. Trist, "The Causal Texture of Organizational Environments," Human Relations, XVIII (December, 1965), 21-32. 70 21. Ibid. 22. Cf. Roszak; Alvin Toffler, Future Shock (New York: Random House, Inc., 1970); Marcus G. Raskin, Being and Doing (New York: Random House, Inc., 1971); James N. Rosenau, Linkage Politics (New York: Free Press, 1969). 23. In the most advanced countries certain parts of the society are already in or approaching an early phase of post-industrialism, while many others remain in various phases of industrialism, and still others are pre-industrial. Under some conditions these different types are interspersed; under others they are separate. With the means of communica tion now available (especially through television) a diffuse consciousness of this total state of affairs is spreading, making reactions more rapid both in unstructured publics and organized interest groups, and altering the threshold of what will be tolerated. Trist, "Transition to Post-Indus trialism. " 24. Emery and Trist, p. 29. 25. Ibid. 26. Trist, "Transition to Post-Industrialism," p. 7. 27. Cf. Alfred Chandler, Jr., Strategy and Structures (Boston: M.I.T. Press, 1962). 28. Gordon Allport, Becoming (New Haven, Conn.: Yale University Press, 1955). 29. Trist, "Transition to Post-Industrialism," p. 4; F. E. Emery, "The Next Thirty Years," Human Relations, XX (Fall, 1967), 199-237. 30. Trist, p. 4. 31. Comments by Chris Argyris, Harvard University, about an earlier draft of this essay pointed out the need to develop this point. I am indebted to him for this and other suggestions. 32. This model was developed earlier and was associated with an argument which attempted to build a theoretical basis for non-bureaucratic organizations. It was developed, in part, because of developments I observed in the organization that is the focus of this overall report and will be described in some detail in the next section of 71 this essay. For additional details and theoretical dis cussion , Cf. Larry Kirkhart, "Public Administration and Selected Developments in Social Science," in Marini, pp. 155- 64. 33. Empirical support for this proposition was provided by a recent study of several industrial organiza tions. Cf. Paul R. Lawrence and Jay W. Lorsch, Organization and Environment (Cambridge, Mass.: Harvard University Press, 1967), pp. 84-158. 34. Neely Gardner, "Organizing for the Future — A Trial." (Paper presented at the 1971 National Conference of the American Society for Public Administration, Los Angeles, Calif., University of Southern California School of Public Administration, May 1971, Mimeographed, 11 pp.) 35. Ibid., p. 1. 36. Professor of Public Administration, University of Southern California. CHAPTER III ORGANIZATION DEVELOPMENT AND ORGANIZATIONAL DEVOLUTION The term "organization development (OD) has been used many times in the preceding test even though it was not rigor ously defined. I shall, following an investigation of formu lations provided by the literature, present a more precise definition of OD in this Chapter. I shall discuss two contemporary theories of OD, Grid OD theory and Situational/Emergent theory, and I shall review research on the consequences of implementing each; I shall explore and advance the strategy of organization devolution; and, show its theoretical congruence to the Situational/Emer gent approach. This discussion will involve detailing the theoretical features of organizational devolution and high lighting its implications when applied to an organization— in this case, the State Compensation Insurance Fund (SCIF). The Concept of Organization Development Precisely when the contemporary meaning of OD came into being is extremely hard to identify. Nevertheless, it is a term which has been used by an increasing number of authors and has appeared in a number of books and articles. Warren Bennis stipulates that 1958 marks the point where organizational 73 change programs shifted from personal growth or personal change to organization development; he says this shift occurred in a program conducted in the Esso Company by Robert Blake and Herbert Shepard.^ This is a convenient point to mark the transition from management development to organizational development. The distinction between these two points of view or strategies of change is extremely important but often poorly understood. Consequently, it is essential that this line be clearly drawn. This division will bestow two advantages: it will enable us to build a conceptual foundation for the further definition of organizational development, and it will permit us to discard from the analysis a large number of articles which would otherwise have to be cited and reviewed, simply because in their usage the shorthand phrase "OD" is not basically different from the older concept of management development. 2 Warner Burke shows m a recent article that OD is directed to changing the culture of an organization— its norms, procedures and general climate. He describes a series of actions or interventions such as team-building and inter-group building, designed to have greater impact than management development efforts. At the same time, management development can be an essential part of OD under circumstances when the management skills to be found in an organization are helping decrease organizational performance. 74 Burke presents a comparative typology that differentiates between OD and management development in six areas: (1) Rea sons for Use, (2) Typical Goals, (3) Interventions for Pro ducing Change, (4) Time Frame, (5) Staff Requirements, and (6) Values. The primary reason for using OD is a need to improve some or all of the systems that constitute the total organization. The main reason for using some form of management development is a need to improve some aspect of the manager.3 Both of these orientations involve training; OD in volves training designed to address a problem which has been diagnosed to be dysfunctional to organizational functioning; management development programs are designed around the needs of managers which have been identified by the consul tant and the managers. In short, OD is focused on system development; management development is focused on the development of individual managers. This distinction makes all the difference between these two approaches to change. Emphasis upon the organiza tion as a whole, as a total system is unquestionably the most important component of the concept of OD. Each of the five following definitions of OD includes the total system idea; 1. OD is a process of continual organizational renewal. The process is one of constantly examining the way the organizational systems are functioning and looking for ways of improving these functions.4 75 2. IOD is] . . .An approach to the management of organizations that recognizes human interpersonal (interactional) process as well as formal organization structure and task content as important and legiti mate matter to be examined; and a. A set of methods to address the above (Behavioral Science Technology) b. An organization ethic that seriously and purposefully attends to the humanness of the people who work within.5 3. ... A realistic behavioral unit of development is found in the organization. The behavioral unit becomes the organization membership including all the significant variables such as policies, rules, regulations, reward and punishment systems, produc tion controls, informal social systems, etc. The organization, then, is the environment within which individuals and groups perform and interrelate toward accomplishing organizational aims. In this sense, the organization is the critical unit of development.6 4. Organization development (OD) is a response to change, a complex educational strategy intended to change the beliefs, attitudes, values and structure of organizations so that they can better adapt to new technologies, markets, and challenges, and the dizzying rate of change itself. Organization development is new and still emerging, only a decade old, so its shape and potentiality are far from granted and its problems far from solved.7 5. Organizational development is an effort (1) planned, (2) organization-wide, and (3) managed from the top', to (4) increase organizational effectiveness and health through (fe) planned interventions In the organization's "processes," using behavioral- science knowledge.8 Definitions 1, 2, 4, and 5 were formulated by OD theorists/consultants who are proponents of the Emergent/ Situational theory; Definition 4 was developed by the persons who invented Grid OD. These definitions encompass the full 76 range of meaning presently attributed to OD. Each takes a somewhat different tack but all either clearly state or imply that OD is concerned with the functioning of the organization as a total system. Beyond this, there seems to be little concensus. To some extent, all the definitions emphasize behavior as the focal point for change. But there is no general agreement on how behavior is to be changed. Because of its vagueness, the first definition has little use in the construction of theory. "Continual organizational renewal" is hardly a concise explanation of the meaning of OD; it encompasses everything, and therefore nothing of theoretical utility. The next definition, number two, is similarly vague and is simultaneously too concrete. The author does not contribute to definitional rigor by conceiving of OD as "an approach to the management of organizations"; this confuses the distinction between management's behavior and the behavior of the organization as a whole; management's contribution is only one part of the overall functioning of an organization. This author seems to place excessive emphasis upon the interpersonal behavior to the exclusion of group and inter-group behavior. The meaning of OD is not clarified by Bennis' definition, number four. He says OD is "a response to change," which he equates with a "complex educational strate gy." This is doubly confusing. Logically, a response to 77 change is a generalized reaction to something and quite different from an educational strategy; these terms could subsume educational strategy plus untold other types of responses. Moreover, "response to change” cannot be used to define a theoretical perspective; it must mean the circum stances associated with the production of theory, rather than a component of the theory itself. Definition number three is a curiosity; several books and articles of Blake and Mouton fail to reveal, with the exception of the quote, anything other than techniques by which they pursue OD: Each time the discussion in the texts seems to be leading to a formulation of OD, emphasis suddenly shifts to the various techniques they developed to intervene in organizational processes. Considering the fact that they are undoubtedly the most famous practitioners of OD and have made major contributions to theory, it is somehow strange they have not conceptualized OD as rigorously as their change techniques. Beckhard's definition (number five) is considerably more concise than any of the others. If one assumes theory and practice should bear some relationship to each other, this definition is undoubtedly right for a large number of organizations. Nevertheless, it is unnecessarily restrictive. Optimally, a definition of OD should be sufficiently general to incorporate all possible forms of organization and mark the boundaries of the field. 78 Beckhard's terminology would restrict OD to organiza tions with traditional hierarchical structures of authority; 9 this is indicated by the terms "managed from the top." This would also restrict OD to a particular type of organiza tional planning process; one which is suited to Industrial and, to some extent, Late Industrial, environments.^ This structure of planning and coordination is not adequate to the complexity of Post-Industrial environments. The crux of the problem with all five definitions is that the authors fail to keep their exposition at the same level of analysis. They shift perspective from the organiza tion as a whole to sub-features of organizations and then equate these two. This can be avoided by maintaining only one level of analysis. The formulation below is intended to surmount this problem and provide a generic definition of OD. In order to do so it is necessary to distinguish organizational development from organizational growth and organizational change. Organizational Growth.— an increase in organizational size, measured by number of people, departments, amount of technology, volume of financial flows or area of society affected by the organization's activities. Organizational Change.— a reconfiguration of tasks, sub- pysterns, influence relationships, budgetary processes or other relationships such that the new configuration can be described as different or changed from what went before but whose effect is not to increase the overall adaptiveness or effectiveness of the organization. 79 Organizational Development.— like organizational change, OD involves a reordering of a socio-technical system from one pattern of relationships and potentialities to another. The crucial factor is that by expending energy to make the change, the possibilities of effective organizational performance are increased. This formulation rests on only two assumptions. An organization can be characterized as a system; and, two, the level at which a system can perform, rather than what it has done, are the definitive measure of effectiveness, regardless of context. This is why the terms "possibilities of effective performance" were used. A system is defined by its internal properties and its relationship with the environment. The two definitions below describe these properties. Internal Properties of a System.— A system is defined as a relationship in which changes in one or more component parts initiate changes in other component parts, and these changes, in turn, produce changes in the parts in which the original changes occurred.11 Properties which Differentiate a System from Its Environ ment.— I I . A system is a bounded region in space and time, within which information and/or energy are ex changed among subsystems in greater quantities and/or at higher rates than the quantities exchanged or rates of exchange with anything outside of the boundary, and within which the subsystems are to some degree inter dependent. 12 Since we are concerned with a particular type of 13 system, that is, socio-technical systems or formal organizations, we are concerned about the extent the design of an organization enables (1) the total organization to draw from the potentiality of its resources and (2) the system to behave as appropriately as possible toward the environment. Daniel Katz, in his article "Group Process and Social 14 Integration," provides an example of organizational growth, perhaps change, but definitely not organization development. He describes the way conflict is managed in a bureaucracy, shows how this involves continual compromise and, accordingly demands continual growth in the size of the organization. Thus he says, "it is much more difficult for a bureaucratic structure to eliminate a subsystem once established than to 15 add two new ones." When the only way a system can move beyond highly destructive conflict patterns is to function on the basis of compromise, thus falling short of meeting the needs and interests of the various subsystems involved in conflict, adding new subsystems is not likely to increase overall effectiveness. It is inherent in the very character of competitive hierarchical bureaucratic organizations that the dominant mode of conflict resolution is through third-party judgement. Two or more parties in conflict are placed in the position of having a third party, generally a superior position in the hierarchy, decide how the conflict is to be resolved.16 This set of interrelationships helps to produce and maintain a continual win-lose orientation between sub-systems. Losers are continually searching for ways to compensate for losses and gain dominance in the future. Winners become overconfi dent about their capacity to affect others; this sets the 81 conditions in which the losers can win more easily. This basic pattern reinforces and deepens itself in a spiral like manner, with the winners and losers changing places as con- 17 flict continues. Growth is promoted and change occurs; development does not occur. Strategies of Organization Development Two distinct strategies of OD can be identified in the literature; both are being applied to organizations with increasing frequency. Grid OD is one of these strategies; it is based on the work of Robert Blake and Jane Mouton, founders of Scientific Methods, Inc. The second, Situational/ Emergent Intervention Theory, is embraced by a much larger number of OD consultants, and cannot be identified with any single person. Grid Organization Development Grid OD has been applied to a number of governmental agencies— the Southwestern Region of the Internal Revenue Service, the Northwestern Region of the Forest Service, a Navy research and development laboratory, and some departments of the National Aeronautics and Space Administration. This orientation is firmly wedded to the literature of the behavioral sciences. Blake and Mouton, in each of their several books and many articles, survey and incorporate an amazing amount of literature. Both are highly regarded behavioral scientists who have been associated with the 82 National Training Laboratories (NTL) for many years. This is especially true of Robert Blake. He has had, and continues to maintain, a close association with NTL and is one of the leaders of a particular school of thought known as the instrumented approach to learning. According to this school, instrumented learning and instrumented training programs have several distinct advantages. They do not require highly sophisticated skills to conduct and can achieve more predictable results than other approaches to learning. It does take a considerable amount of sophistication to create the initial design. Goals of the program and the most important types of behavior to incorporate in the design must be utterly clear. Each component in the program must be prepared in such a way that it evokes the type of behaviors that must be forthcoming in order for the learning goals to be achieved; this requires the program designer to walk an extremely thin line between too much specificity of the roles various program participants are asked to take and too little. If the training roles are too specific, there will be little likelihood that the trainee will behave in terms of how he naturally acts; instead, all that will be forthcoming is the role he is required to take. If the roles are not specific enough, behaviors unrelated to the learning goals may be evoked; if this occurs, the outcome of training becomes unpredictable. 83 A program of this kind must necessarily be developed beforehand: assumptions must be made about the persons who will receive the training. Since it is not generally possible to base the program on knowledge of the particular individuals who will be involved in the training, instrumented learning must rest upon a different source of knowledge. This source is behavioral science. Drawing from this litera ture, programs are developed to address general aspects of human behavior. Cogency of a program to any particular organizational problem or human situation depends on how general the problem or situation is in society. And, it depends on the degree to which the underlying behavioral research has produced reliable and valid knowledge of human 18 behavior of a particular kind. Instrumented learning is a distinctive feature of Grid OD; it consists of a packaged program of instrumented training which proceeds sequentially through six specific stages. Phase 1 of the six-phase approach involves studying the Managerial Grid as a theoretical framework for under standing behavior dynamics of the corporation's culture. In Phase 2 the behavior dynamics of actual organizational teamwork are studied and tested in settings of actual work against the Grid model for the perfection of problem solving methods. The same kind of application is made in Phase 3, but to the interworkings between organized units of the company where cooperation and coordination are vital to success. The top team in Phase 4 engages in a study of the properties of an Ideal Strategic Corporate Model necessary to bring corporate profit ability logic to a maximum-thrust condition. Phase 5 involves implementation tactics for converting the corporation from what it has been to what it will become 84 under the Ideal Strategic Corporate Model. Phase 6 measures changes in conditions from pre-Phase 1 to1 post- Phase 5 for the evaluation and stabilization of achieve ment and for the setting of new goals and objectives of accomplishment for the future.I9 With the exception of the first stages, the Grid program is administered by the organization with minimal assistance from an outside consultant. Self-administration is possible because each Phase has been very thoroughly instrumented. In each of these programs the members of the organization are required to participate in small group, experiential exercises; when the exercise is over, the trainees complete a paper and pencil test. This test is designed to enable the participant to describe and assess his or her own behavior with a relatively short questionnaire composed of "forced choice" items. The various items are scored in a special way to yield a generalized description of the individual's behavior. Each of the several generalized descriptions that can be produced are representative of a particular class of organizational behavior described by Blake and Mouton in their theoretical works. After each trainee has completed this paper and pencil assignment, he or she shares his or her "self- description" with the other members of the group. The other members' perception of this person's performance is discussed and differences between the individual's description and the group's are explored. 85 Theoretically, this strategy takes three to five years to be fully implemented; today, very few organizations, if any, have moved beyond Phase 3, the intergroup relations program. That is, itself, a rather interesting fact, which probably is in part a testimony about how difficult it is to effect change in an organizational culture. This strategy prescribes an ideal, desirable end- state for an organization to achieve and it is, very clearly, 20 the best-articulated OD strategy presently available. The basic proposition of Grid OD is the most effective way to manage any organization is to adopt a style which fully integrates both concern for production and concern for people and uses work teams as the basic building blocks of 21 the organization. The way the program begins is to have a number of members of top management attend a six-day Managerial Grid training program away from their organization; the training design is based upon techniques Blake and Mouton believe are most effective in changing individual behavior. In contrast to a non-instrumented learning laboratory, where the basic data-gathering process is undertaken by the members of the training group and worked in the "here and now" as the members feel it is necessary to do so, the instrumented learning lab used by Blake and Mouton has a set of prescribed events and exercises which shape the direction and development of the group and its participants. 86 Trainees are provided with a well-articulated descrip tion of their behavior by other members of the group involved in the problem-solving exercises; the parameters of this description are drawn from the book The Managerial Grid and designed to show an individual the managerial style he is currently using and what behavioral skills he needs to develop in order to move more closely to the optimal "9,9" style which maximizes concern for production (or task) and concern for people. This not only has the effect of generating strong forces for individual change, it also means that by going through this process, the members of the group are likely to adopt the Grid value framework. The overall effect of this design is, therefore, to set in motion sociological forces which ultimately could begin to change the culture of the target organization. Following this session, additional members of the target organization are trained to run the exercises; they can return and train other members of the organization. This is the usual set of events that occurs before Phase 2, team training, is launched. The team training program is designed to be applied to organizational networks of superior and subordinates who work together from day to day. Once this phase has been implemented widely in the organization, the third phase begins. This program also centers on teams but rather than focus on developing work units, it is designed to 87 build more effective inter-group relations within the organi- 22 zation. Again, the cultural norms of the organization are the targets. In summary, this strategy is designed to begin with top management, train it in the values and application of the Grid system, train organization members to conduct the Grid programs with on-site work teams and build more effective inter-group relationships within the organization. The last three phases involve the development of an ideal corporate model by top management, its implementation and, finally stabilization of the organizational culture around the fea tures of the ideal model. Grid Organization Development Applications Three studies are available that document some of the effects of Grid OD. The first of these explores the results of applying the first two phases to an industrial organization 23 called Sigma Corporation in 1962-1963. The second study documents the results of undertaking Grid OD in several 24 regions of the U.S. Forest Service. Like the previous study, this one is limited to the effects of the first two phases. It was conducted in 1965 and 1966. The last study reports the consequences of implementing the first two phases to several segments of the Internal Revenue Service during 1964-1966.25 88 Sigma and Grid OD Results of the Sigma program are summarized by two 26 articles, "Breakthrough in Organization Development" and 27 "A Second Breakthrough in Organization Development." The first article has received considerable attention; the second is considerably less well-known. Bluntly, neither of these studies provides a solid basis for knowing the results of Grid OD. For two different and equally serious reasons, the first of these studies does not provide the reader with anything more than a tentative and highly speculative conclusion that the Grid may have had some effect upon the attitudes of managers; despite the authors' arguments and flights of fancy, there is very, very little reason to believe the program had any effect on the overall performance of the organization. Comments by the authors themselves are sufficient to show one of the reasons the OD evaluation cannot be taken seriously: There were significant increases in productivity and profits during 1963, when the organization development program was in effect .... Total production rose somewhat (with fewer employees), and profits more than doubled. At first glance, it would seem that Sigma had struck gold, that its worries were over, and that the development program had been highly effective. But this in itself would be a gross oversimplification. To begin with, Sigma's business involves widely fluctuating market prices, raw-material costs, and other non-controllable factors. Possibly higher revenues or lower materials costs would explain profit increases. 89 In addition, new automatic machinery and new plant equipment investments might be sufficient cause for the reduced labor force and increased profit picture. Finally, an overall manpower reduction had occurred (involving over 600 employees), and this in itself might account for the increased profit picture in 1963, particularly if the increased overtime costs (at time- and-a-half) had been spread over the remaining work force. These possibilities make it difficult to draw simple cause-effect conclusions about Sigma's develop ment program and operating performance.28 Virtually every possible factor that could compound the problem of assessing the OD program can be found in the previous paragraph. The authors explore a number of ways to sort out the effects of these various factors but their propositions would not convince even a moderately skeptical mind. The other equally serious problem is the research design employed by the study. Since it does not involve a control group there is simply no way to know if the changes in Sigma Corporation were unique to it or were the result of widespread phenomena that affected all organizations similar to Sigma. Historical developments are a plausible explanation of Sigma's remarkable profitability; given the research design, there is simply no way to refute this 29 possibility. And, since the research team did not gather data before the Grid program was implemented, there is no respectable way to assess change in attitudes or other non economic measures. 90 The second article provides even less useful informa tion than the first. It is simply a discussion that offers no evidence; apparently, it is grounded upon a similar research design. The article celebrates the fact that the Grid program was extended to the lower level, non-managerial employees and provides anecdotal information.'*® Despite the general reputation of this study, nothing of any significance can be learned about Grid OD through these two articles. The Forest Service and Grid OD The Forest Service study is vastly more sophisticated than the ones in the articles previously surveyed. Greiner was involved in both the Sigma study and the Forest Service study; apparently the decision to evaluate the results of the Sigma project was made after the training was well underway; in the case of the Forest Service, the decision was made at an earlier date. Consequently, it was possible to develop a much more elaborate research design and strategy for measuring results. Data gathering began in December, 1965, and ended in December, 1966.31 A number of different features of organizational performance were studied; they included (1) organization performance (managerial evaluations, output, costs), (2) behavior (boss-subordinate relations, work group relations, individual job practices), (3) attitudes 91 (satisfaction, recognition, loyalty) and (4) values (work group norms). For some unexplained reason, the first block of data consisted of attitudes toward Grid training after the first phase. These data were collected at the conclusion of each of the one-week Grid seminars. Predictably, these data showed that the participants regarded the training as an 32 important learning experience. Anyone who has ever seen training evaluations of this type— immediate post-reactions— knows it is extremely rare, especially if the program lasted three days or more, for participants to rate the value of the training low. By far the most common practice is a very favorable rating— such as occurred in the Forest Service. These data are not, therefore, very useful indicators or predictors of how the training will affect behavior back in the organization. The "environmental press" a person faces in a work situation is markedly different from what he or she faces in the con cluding moments of a training program.^ While these data may be helpful in determining whether the training program achieved its learning goals, they are of little relevance compared to on-the-job measures. In addition, there is no basis of comparison involved in the post-training data because there were no control groups. 92 The balance of the material in the study rests on a much firmer foundation. The research design incorporates three experimental groups— 184 persons from Region 1 of the Forest Service, 112 from Region 8 and 157 from the Washington headquarters— and four control groups. One hundred fifty- five persons served as a control group for Region 1, 113 persons who did not receive training served as a control group in Region 8, 20 people from the Washington headquarters served as a control group, and, finally, 99 people from Region 2, a segment of the Forest Service that was not in any way exposed to Grid training, provided additional base line data. Two months after the first phase of training had been completed, Greiner administered a questionnaire to the trainees which was designed to assess the impact the program had on job behavior. He reports that significantly more managers saw their bosses changing their behavior than did the control populations that did not receive training. The worth of this finding is doubtful; it is based on responses to one item taken from the questionnaire. This is not a highly reliable method for obtaining information. It is generally conceded that a much more reliable way to measure reactions to a particular phenomenon is to use a number of items designed to assess various facets of the phenomenon and yield a scale score. 93 In addition, the item is itself rather vaguely worded: "To what extent has your boss applied training ideas 34 in the past two months?" A third block of data was gathered six months after the initial training program. By this time, some of the trainees had gone through the second phase of the Grid (work team improvement). Using the same item discussed previously, Greiner found that managers who had gone through only the first phase were not significantly different from the managers who did not receive any training. Managers who received training in both the first and second phases scored significantly higher on this measure than either of the other groups. Greiner points out that these findings show that the impact of Phase 1 training had faded out six months after the program. Compared to the measurement achieved two months after the training, the managers who had completed both phases did not produce a markedly higher average score on the questionnaire item. The response options for the item ran from 1 to 7, with 1 representing 35 "very infrequently" and 7 representing "very frequently." The average score on this item was 4.13 two months after beginning training; average score for participants in both phases four months later was 4.48. To the extent one can believe this result, it suggests that unless the first phase is quickly followed by the second, the effects of 94 training will be erased. Greiner also shows that the second phase did not strengthen managers' acceptance of Grid 36 theory. The chief reason why it did not is that they found it difficult to use the theory when their subordinates 37 had not received similar training. Finally, an elaborate evaluation effort called the 38 "judges project" was employed to assess the effects of Grid training on organizational performance. Heretofore, the study populations have not been differentiated; for this analysis the populations investigated were broken down into regions, forests and districts. In relation to the regions, the author says: . . . The Grid program possibly contributed to improved overall regional performance. This is the most positive indication of Grid program effects on performance, since forests and ranger districts do not show similar improvements.39 Based on the material produced by the "judges project," the tentativeness of the author's conclusion seems fully warranted. The project results took the form of overall performance rankings and were displayed by Greiner with bar graphs which could easily be interpreted as showing no effects.^ Internal Revenue Service and Grid OD Although relatively short and direct, the IRS study is considerably more rigorous than either of the two previous research efforts. As was the case in the earlier studies, 95 Larry Greiner was one of the researchers. The study is focused on the consequences of experimenting with Grid OD in one of the large Regions of IRS. The Regional Commissioner got training in the first phase, started in two district offices in 1964. In 1965 two additional offices underwent that training and in 1966 two 41 more were involved. Although the timing of the second phase of training is not explained, the authors indicate this program had been completed before 1967 and that the third phase was beginning. The research design involved eight district offices. Six of the district offices received training during the period of research and two did not. The training program was applied sequentially; two offices were trained in the year preceding the research, two were trained in the follow ing year and two more the next. All but two of these offices had worked under the direction of the Regional Commissioner. The reader is not told whether the two offices that did not receive training worked under the same commissioner or for another. This is quite clearly the strongest of the research designs that have been discussed. As was the case in other studies, data were gathered with a questionnaire. This particular questionnaire seems to have been especially well designed and is far superior 42 to the instruments used in the other studies. Any criticisms of the research instrument would have to be 96 directed at the methodology of behavioral science in general, not to this particular questionnaire; it was developed through exceptionally rigorous methods. The questionnaire measured organizational climatei Organizational climate refers to the ways in whlon an organization is sees as fitting together task and people inputs in handling its key problem areas. (Italics in original)43 Based on the theory of the Managerial Grid/ discussed earlier, five types of organizational climates were conceived: (1) Directing, (2) Integrating, (3) Withdrawing, (4) Compromising, and (5) Smoothing. The authors predicted that the implemen tation of the Grid program would produce greater emphases on Integrating (high levels of concern for production and for people) and that this change would occur and be greatest during the second year of the program. The authors reached this basic conclusion almost without exception in every area they explored: Our findings revealed very little change in the climates of the eight districts over a three- year period. Moreover, those climate changes which did take place could not be attributed to the Grid program.44 Interpretation of the research data was unavoidably and surprisingly consistent. Only in one or two cases were any changes in climate found and these occurred simultaneously in both the experimental and the control districts. 97 The authors concluded that their findings were largely attributable to the role of the National IRS head quarters, which dictated the operations of the various parts of the organization to a very high degree. They point out that "...finding new ways to reconcile the need for consis tency on some issues while promoting flexibility around other issues,"45 is a dilemma faced by many modern, complex organizations; it follows that the effectiveness of Grid OD is to a large extent determined by this dilemma. Summary I have explored the theoretical features of Grid OD, paid particular attention to the instrumented training design that it employs and have examined the research findings from three different studies of the application of Grid OD. The first of these studies was so inadequately designed and subject to extraneous and compounding factors that I did not regard it as useful in understanding the potential of this strategy of change. The second study, which was undertaken immediately after the first, was based on a better research design and utilized some measurement techniques that were well-conceived and some that were not. Overall, this study showed that the first two phases of Grid OD had little effect upon the operations and attitudes of the focal organi zation, the Forest Service. The study showed that unless the second phase follows shortly after the first, the limited 98 gains achieved in the first phase will be lost. The third study was superior in every way to the previous two. It employed an excellent research design and used superior methods. Still, however, the Grid program did not produce any measureable effect upon six district offices of the Internal Revenue Service. Emergent/Situational Intervention Theory Although the general features of what I have been calling Emergent/Situational Intervention theory can be des cribed, it is remarkably difficult to specify exactly how the theory evolved to its present form. A considerable amount of the components of the theory have their roots in the early experience of the National Training Laboratories beginning as early as the late 1940's. This organization was heavily involved with the development of the laboratory form of training, especially in terms of management develop ment programs. In an article which describes the evolution of T-groups (training groups) during the last 20 years, Kenneth Benne points to events in New Britain, Connecticut during the summer of 1946 as giving rise to the particular 47 orientation that was later called the laboratory method. To some extent, this is disputed. Carl Rogers says that comparable events were occurring at the University of 48 Chicago at almost exactly the same time. It appears that this may be a case of simultaneity of invention. For our 99 purposes, the most important concerns are the characteris tics of this orientation. The laboratory method, as it was applied to management development was meant to be extended to organizational change, had two goals: "(1) to effect a change in values so that human factors and feelings come to be considered as legitimate and (2) to assist in developing skills among 49 managers in order to increase interpersonal competence." An intentionally simplified and brief description of the lab oratory method is provided by Warren Bennis: It unfolds in an unstructured group setting where the participants examine their interper sonal relationships. The training process relies primarily and almost exclusively on the behavior experienced by the participants; i.e., the group itself becomes the focus of inquiry. Conditions are promoted whereby group members by examining data generated by themselves, attempt to under stand the dynamics of group behavior, e.g., deci sion processes, leadership and influence processes, norms, roles, communication distortions, and effects of authority on a number of behavioral patterns, personality and coping mechanisms, etc. In short, the participants learn to become more sensitive to the processes of human interaction and acquire.concepts to order and control these phenomena. Although the phrase "here and now" has become a cliche in recent years, it aptly captures one of the import ant aspects of the method, its emphasis on present experience. This aspect was and still is for the most part promoted by the trainer's role in the early moments of the program. The trainer, instead of defining what the training group should think about or believe, assumes what is by conventional standards, a novel role. He or she repeatedly refuses to assume the traditional leadership role and tell the group what it should do; instead, he or she emphasizes how the group is behaving at the moment. By refusing to assume responsibility for the group, the consultant provides a role-model that is crucial to the effectiveness of this method. He assumes responsibility only for himself. He frequently intervenes in the group's processes when inter personal objectification occurs; i.e., when one or more persons are telling one or more other members of the group "what they are" rather than sharing how they are reacting to the other's behavior. As the group members discover that the former process usually prevents change and the latter promotes it, the predominant motif of dialogue shifts from remote experiences in vague and unspecified circumstances to more and more immediate experience in the "here and now" situation of the group. This, in turn, enables the group members to become more aware of and sensitive to the com plexities of events that unfold within the group and to 51 deal with this information more effectively. To the extent these events and processes become vital to the group, basic redefinitions of the potential of human interaction occurs. When individuals interact while playing roles, as is perhaps most frequently the case in social 101 relationships both outside the laboratory and typically in the early moments of the training program, interpersonal problems are more rarely confronted directly compared to what can be achieved when interaction occurs at a more personal level. Initially, this method was used for executive or management development. It was assumed that a manager learning more about himself and group behavior would promote organizational change in that context. The overwhelming weakness of this strategy was made very clear by the famous International Harvester study of the early 1950's. This highly sophisticated and well-designed study revealed that T-group training could and did change behavior of individual managers (as reflected by pre- and immediate post-training measurement); but what most deter mined the application of this learning on the job was the behavior of the immediate supervisor. If he supported his subordinate, acting on what he learned, that learning did indeed find expression in the job; otherwise the subordinate 52 tended to behave as his supervisor did. These research findings played an important role in shaping the Emergent/Situational Intervention theory. In creasing attention was given to training groups composed only of persons who worked in the same organization, and to applying the approach I have been describing to problems 102 53 that the organization and its members were facing. Meanwhile, new training designs were invented and new theories were put forward. These events were taking place in the late 1950's and early 1960's and were signaled 54 by increasing emphasis upon team-building laboratories, planned organizational change,55 and the emergence and grow ing use of the term "organizational development."56 Although no consensus was achieved on the name to be given the role played by the consultant (today, there is still no consensus and the terms "OD consultant", "change agent", "process con sultant", and "organizational change specialist" are used interchangeably), the basic aspects of the role were increas ingly well defined. In a recent publication, Edgar Schein calls this role "process consultation" and characterizes it this way: It is a key assumption underlying P-C [process consultation] that the client must learn to see the problem for himself, to share in the diagnosis and to be actively involved in generating a remedy. The process consultant may play a key role in helping to sharpen the diagnosis and in providing alternative remedies which may not have occurred to the client. But he encourages the client to make the ultimate decision as to what remedy to apply. Again, the consultant does this on the assumption that if he teaches the client to diagnose and remedy situations, problems will be solved more perman ently and the client will be able to solve new problems as they arise.57 103 In contrast to Grid 0Dr which was described earlier, Emergent/Situational Theory places a great deal of respon sibility upon the consultant in OD programs. To be effective, the consultant must know himself, be highly skilled in inter personal interaction, be at least aware of the basic theories of individual, interpersonal, group, and intergroup behavior and, finally, be adept at applying existing training designs and creating new ones which are appropriate to the organiza tional situation. No specific interventions are planned a priori for an organization undertaking or about to undertake an OD effort. Both Emergent/Sj.tuational theorists and Grid theorists share the belief that it takes two to five years of concerned OD efforts to change the culture of an organiza tion. Intervention Techniques For an Emergent/Situational (E/S) consultant, there is no single best way to intervene in the process of an organization. The test of a "good" intervention is the ex tent to which it is based on the organizational situation and furthers its ability to solve its own problems. At the same time, there is a set of conditions which E/S theory suggests must exist before it is possible to intervene effectively. These conditions have been described in detail 104 in a recent book by Chris Argyris called Intervention Theory and Method. The first of these is valid information shared among the members of the client group or organization and between the consultant and the client. The second condition is free, informed choice; the intervention "... should be so designed and executed that the client system maintains its 58 discreteness and autonomy." Finally, "...the third basic process for any intervention activity is ... the client's internal commitment to the choices made."^ Given this orientation, there are a number of distinct techniques for change that can be utilized.60 One is team building which involves working with an organizational family (superior and subordinates) to build more effective relationships; a consultant interviews the various members of the family prior to the laboratory session, classifies the problems that have been revealed during the interviews, feeds this information back to the entire group at or near the beginning of the session, and then helps the group work on the identified areas of difficulty. Unlike the design used by Blake and Mouton, this is not an instrumented learn ing situation. The issues which are worked on during the training are those which were identified by the client during the pre-session interviews or during the session itself. A second intervention is similar to the Blake and Mouton intergroup program but follows a less structured pattern. Typically, the consultant employs this technique when he discovers that intersystem relationships are the factor reducing the performance potential of part of an organization. Although the specific applications vary, generally the procedure is to identify the several persons most crucial to each group's ability to change its relation ships with another group and then to bring together the two sets of crucial persons. The process is begun by asking each group to share its image of the other group; the represen tatives of the other group are asked to make sure they under stand as completely as possible how the other group perceives their behavior without evaluating what the representatives of the other group have said. After this, the second group tells the first group how it is perceived. Thus the groups clarify where they stand with each other. Ultimately, the groups begin, with the help of the OD consultant, to work on their relationships. A third technique, Deep Sensing, involves bringing together a cross section of organizational members to talk to an organizational leader about how they feel about the overall policies, pay plan, promotional opportunities and general climate of the organization. The manager's role is to make an all-out effort to understand how this representative group sees the organization and not defend either himself or existing practices. Once he understands this, he tells 106 appropriate persons in the organization and necessary adaptations are made. It goes almost without saying that in order for this technique to work there must be a great deal of trust between the representative group and the leader. A fourth technique is known as the "organizational mirror". Data about the performance of an organizational subsystem are gathered from the environment. More often than not, they are perceptions of persons of groups who interact with the unit on a regular basis. In some cases the kind of data sought will be specific and in others general infor mation related to anything the people in the environment of the system want to let the system know. At any rate, after the information is gathered, it is fed back to the subsystem and whatever action is called for is taken. This list could be extended; however, the point is that the E/S approach to organizational development is based upon the premise one must begin with the system "where it is" and avoid activities defined a priori. The focus is on existing problems which are sufficiently pressing that per sons in the organization are willing to work on them. Having discussed the E/S theory and techniques in some detail, we are now in a position to examine several important research efforts that explore the effectiveness of this approach. 107 Applications of Emergent/Situational Theory The research material reviewed in this section will present findings on the effects of E/S organization develop ment in three different organizations. The first program to be examined was conducted with several divisions of a bank located in New York City; the second study is based on a large research and engineering company; the third is an exceptionally long-term study of organization development in a garment manufacturing firm located in Pennsylvania. The criteria used to select the studies were: (1) the program was oriented to the organization as a whole, (2) the program involved a serious commitment to OD that extended for more than two years, and (3) empirical measures of the effects of the program were obtained by the investigator. Although not previously stated explicitly, these are the same criteria used to select the studies reported in relation to Grid OD. Organization Development in a Bank This study describes a series of interventions made in the organizational processes of a large commercial bank.61 The program was primarily concerned with creating more effective relationships between two highly interdependent departments. The program lasted for slightly more than a year, including the time spent by the consultants before any interventions. For all practical purposes, this program is an 108 example of intergroup building. The research design incorporated an experimental unit and, for the most part two control groups. Data were collected immediately before the OD program began, during, and immediately after. Four years later, an additional block of data was added in order to assess the long-term effects of the program. The research monograph is somewhat uneven; it contains a considerable amount of anecdotal material which will not be reported, several reasonable well-designed attitudinal measurements and routine performance data gathered by the staff of the bank. An exceptionally long list of circumstances had to be explored and discussed by the consultants and top manage ment of the bank before the program began; eighteen months elapsed before top management and the OD consultants started the experimental intergroup program with the mortgage produc tion group and the methods group. A special experimental team composed of nine managers from the mortgage group and three from the methods group were involved in the program for the "...purposes of improving their operational effectiveness as a problem-solving unit and facilitating the introduction of 6 2 the computer into mortgage production." Each of the members of the team participated in a sensitivity training program immediately prior to the series of three-day conferences which were the heart of the OD effort. 109 Just before the first of these conferences, the consultant collected data from the members of the team on obstacles to the effective operation of the team; this information was used in the first three-day conference. Six months later a second conference was held; "the conference was planned by the team and included an agenda of priority problems that they be lieved required more time than was available in their monthly 6 3 meetings." Finally, the team met for another three days approximately six months later and decided there was no need for further assistance from the consultant. The anecdotal material gathered from interviews with various people around the experimental unit seemed to indicate that the program had positive effects. Since the methodology behind these interviews and the decisions used to include or exclude material are not clear, any conclusions based on this material are more or less an act of faith on the part of the reader. Routine measurements of organizational performance seemed to indicate that the productivity of the experimental unit increased as a consequence of the program but not by much. The program did have a strong, positive effect on turnover; when the program began the turnover in the experi mental unit was considerably higher than that of either of the control groups; at the end of a year, turnover was down 110 in mortgage production and was increasing in the control groups. Absenteeism also was significantly reduced as a result of the program. The program also increased communication between superiors and subordinates in the experimental unit and the perceived influence of the managers of the experimental unit on their superiors relative to the control groups. Many organizational innovations, programs and special relation ships were attributed to the OD effort. Shortly after the initial one-year study was com pleted, a number of team-building efforts were undertaken with other units of the organization; these efforts pro duced considerable change in the structure of the organi zation. At the same time, these changes made the follow-up study conducted four years later because of the many changes in the intervening period. Thus, despite efforts to create "comparability" of the general measures used in the overall study, the value of the last block of data is essentially that of an assessment of the overall organization at that point in time. OD in a Research and Engineering Company This report is based on a large and elaborate program 64 conducted in a highly professionalized organization. Prior to the decision to start the OD program all vice presidents of the company attended a T-group. Shortly thereafter, a Ill series of six "cousins laboratories" was conducted with sub ordinate managers two levels below the vice presidents; the labs were spread over a year period and each lasted five days. Concurrently, a series of two-day, team-building sessions was undertaken with groups of division directors, the vice presidents and the president. The research design of the study was a conventional experiment-control group configuration. The population in volved in the experimental group ranged from 56 to 90; there were 76 subjects in the control group.Research data were gathered three times in the course of a year; the first bloc was obtained before the program (the cousins labs) started; the second was gathered at the training site immediately after the lab was completed, and the third block was acquired a year after the beginning of the program. Since the second measure ment is subject to the same criticisms levied during an analysis of one of the Grid OD studies, this material will not be reported. Comparative analysis of the research data gathered from the experimental and the control groups prior to the program revealed that the two groups were closely matched; one year later there were significant differences between the two groups. The experimental group perceived a signi ficant increase in the extent to which people in the organi zation openly faced conflict and the experimental group per ceived an increase in the willingness of managers to ask for help. Between the beginning and end of the study the organi zation underwent a major reorganization. Data were gathered about the extent to which the new organizational structure was appropriate for the functions of the experimental and non-experimental groups. Members of the experimental group were significantly more confident to the extent that they believed the new structure was adequate for their tasks and less convinced than the non-experimental group of the appro priateness of the structure for the general functions of the organization. The researchers attribute the first of these findings to the increased confidence that the experimental group had in the possibility of working with other segments of the organization and facing conflicts. They interpret the second finding to mean that the non-experimental group believed the structure was more adequate simply because it would not force as many conflicts into the open; the experimental group thought it less adequate because it would not bring more conflicts into the open. Two years after the beginning of the program the researchers organized a special project to determine if the managers who were highly active and aggressive in the labor atory exercises were also activists and innovators in their jobs. The study supported this hypothesis. 113 : Finally, data were gathered on the managerial values embraced by the experimental and control groups before and after the OD effort. It was found that there were no differences between the groups prior to training and none a year after training. This program seems to have increased the ability of the organization to handle conflict, to have promoted more risk-taking behavior and to have had no impact upon ex pressed managerial values. OD in the Garment Industry Finding a research project in the social sciences, especially in OD, that covers more than a few years is rare; the study about to be reported covers a seven year period and provides what is undoubtedly the most elaborate and compre hensive research findings that have appeared in the literature 66 on organizational change. The study is extremely complex because of the nature of the change effort and the circumstances in which it took place. The organization (Weldon Co.) was a competitor of another garment manufacturing firm that became famous in the early 1950's by its experiments in participative management. Early in the 1960's, ownership of the firm shifted to the second organization (Harwood Co.). When this occurred a study was made of the newly acquired organization and it was found that the organization had decayed almost to the point of com 114 plete malfunctioning; this condition was attributed to poor planning, poor personnel practices, union management relations, high employee turnover and absenteeism. Management of Harwood Co. decided to undertake a major effort to rebuild Weldon Co. New technology was introduced; the work flow was changed and new production controls were developed. An elaborate organizational development program involving all levels of the organization was started. In many ways the organizational change program was not representative of Emergent/Situational theory. For example, team building sessions were started with top management of Weldon Co.; these sessions were unusual because they were marathons that ran for two days and participation was manda tory rather than voluntary; the latter is contrary to one of the basic features of E/S theory — free choice. Moreover, the description of the overall change program gives the impression that Harwood management was taking no chances; every feature it believed an organization should have was incorporated. For these and other reasons I have classified this change program under the heading E/S approach even though it is not an example. I did so for want of having any other classification and inability to discard it from the presentation. A more apt characterization would be "planned organizational change: a prescription for participative management". 115 The change program had three main defining features: we wanted to retain the existing Weldon personnel if this could be done; we wanted to modernize the physical plant and work methods; we wanted to introduce a new pattern of organi zational life emphasizing participative leadership and management. Also, we wanted to accomplish these changes as soon as possible.6? Considerable information could be reported about the results of the change program; since almost all of the materi al discussed in the previous studies involved attitudinal data, summary findings in this area are all that will be re ported. The findings are related to: 1) what people wanted from a job, 2) opinions about the company, 3) opinions about their own job, 4) compensation, 5) the future and 6) per ceptions of change. A series of questionnaire items were used to provide information in each of these areas. All findings are based on a comparative analysis of Harwood Co. 68 (the control group) and Weldon Co. (the experimental group). The organizational change program undertaken by Weldon did not have significant positive effects upon employee attitudes toward the company; employees of both organizations were very positive about their jobs and not significantly different from each other in their appraisal; two years later nothing had changed. In 1962, employees of Harwood reported that they had a greater work load than did employees of Weldon. By 1964, both organizations were reporting similar work loads. This was a significant change and indicates that the Weldon program increased the work required from its 116 employees. As a consequence of the change effort, Weldon employees thought their pay improved; the attitudes of both the experimental and control groups improved during the two- year interval between data-gathering efforts but the improve ment of Weldon employees was more marked. Attitudes toward fellow employees were generally positive in 1962 and had not changed as of 1964. One of the most prominent findings of the research was the sharp change that occurred in career plans. When the first block of data was obtained, a large proportion of Weldon employees revealed that they were seriously contem plating leaving the organization; when the second block of data was examined, the researchers found that there had been a very sharp decline in the number of Weldon employees who contemplated leaving. The next finding is very curious. Despite what would seem to have been major changes in supervisory practices, technology, work load, etc., employees seemed to be unaware of the changes. The Weldon employees, with the exception of a small minority, seemed somehow to have forgotten that major features of their jobs had been changed. In summary, the authors suggest that: ...basic gains in the "output" of an organi zation with respect to satisfactions, motivations, and positive feelings often may be harder to achieve than gains in cost performance and work output."69 117 Several aspects of this study and the change program stand out. Chief among them is that the program did not involve any ideas which had not been thoroughly tested. In fact, the participative management approach had its beginnings in Harwood Company's experiments about ten years earlier. A second aspect is that change did not occur because top managers of Weldon wanted it; they were required to involve themselves in several training programs. These programs were expected to change their attitudes toward the need for change; and, following the programs, their attitudes seemingly did change. Was this because of the training program or was it because of recognition of the inevitable? Finally, despite the enormous changes in Weldon, its employees remained relatively unaffected. The reader wonders whether this was not a consequence of the nature of the change program. Ordering employees to perform in a certain way without consultation is essentially the same as deciding that employees will work in a "participative" organizational structure. From the viewpoint of an employee neither of these options is one which would affect the nature of his in volvement in the job; in neither case does the employee have a real choice. The key ideas about organizational structure and organizational process embodied in the Weldon program are not really very "advanced" or "experi mental" but merely well chosen and applied.70 118 OP Strategies: Retrospect and Prospect I have considered two contemporary strategies for organization development from the viewpoint of (1) their theory and premises and (2) in terms of the consequences of applying the theories to organizations. Research studies associated with each of these approaches ranged from very well designed and executed (Greiner's IRS study) to very poorly designed and executed. The studies associated with Grid OD seemed to show that the more effectively the research was pursued, the less effective the Grid program appeared. Findings seemed to be somewhat more positive when it came to the application of Situational/Emergent Theory; al though the study of the bank and, to a lesser extent, the study of the engineering firm were not as neatly designed as the IRS study, they were respectable research efforts. In both of these studies more change occurred as a consequence of the S/E approach that was the case with the Grid approach. This generalization must be viewed as highly tentative. Finally, I reviewed a program for changes reported along with the S/E research studies. It appeared at this point in the text simply as a matter of convenience, not because it represented an application of S/E theory. The notable feature of all of these studies is the extent to which nothing novel occurred. One or two of the OD programs seemed to have affected the structure of the 119 organizations but not in a way that produced any new knowledge about the possibilities of organizational design. Since one of the chief concerns of OD is improving the performance of the organization as a whole, the lack of interventions in organizational processes which were directly intended to affect the total organization is notable. For Grid OD this is not surprising; Blake and Mouton's program is not supposed to deal with this question until the fourth phase has been reached; in the reported studies none of the organi zations had gone behind the second phase. And very little was said about the organization's environment — either in the goals pursued by the OD programs or in the research monographs. This suggests, to the extent that this is representative of what is happening in OD in general and to the extent that the environments in which the OD programs are being pursued are turbulent, the need for novel OD intervention exists. The intent of such interventions should be to promote increased organizational adaptability and effectiveness. In the next section I shall present one example of an OD intervention designed to effect the total system. 120 The Strategy of Organization Devolution At the beginning of this essay, I devoted several pages to defining the term "organization development". Con ceiving an organization as a system was an idea common to the definitions of OD that I reviewed and to the one I proposed. I defined organization development as a particular type of organizational change which increased the capacity of an organization to relate effectively to its environment. Several of the terms employed by this definition — capacity, relate, environment — are associated with systems theory. Since the theory of organization has long been con cerned with such issues as performance, efficiency, control, and coordination, it is not accidental that modern systems theory and the terminology associated with it have been absorbed into this theory. This approach is especially useful as a tool to explain the concept of organizational devolution. Systems theory is essentially a way of thinking that calls attention to the behavior of wholes. Relational terms such as boundary, independence, dependence, and interdepen dence are basic tools in understanding the behavior of wholes; analyses of the relationships of a part to the whole and the whole a part (sub-system to system and system to sub-system) are additional tools to help create understanding. The first attempt to formulate the theoretical under pinnings of the concept of organizational devolution can be 121 found in an unpublished article by Frank Sherwood called "De- 71 volution as a Problem of Organization Theory". The essay is concerned with a specific type of relationship of sub system to system; the role of local government in the national political system. He suggests that this relationship should be construed as a problem of organizational strategy. By defining the problem in this manner, he was led to consider the ways in which the relationship between local government and the national political system could be conceived. He shows that almost all theories about this relation- 72 ship have been in terms of hierarchy. That is, a relation ship in which one sub-system dominates another sub-system through the exercise of power and authority; conventionally, this is visualized as a pyramid of power in which the compon ents at the bottom have the least power and those at the top the most. A common way of conceiving the role of local govern ment is to subordinate it to the national government. When local government is highly subordinated to national government a political system is said to be centralized. If the relation ship is reversed, and local government dominates the national government, the political system is decentralized. The common factor to both of these types of political systems is a hierarchical relationship between the components. Thinking along these lines has led to the idea of a centralization- decentralization continum. 122 The continuum assumes that there is a certain quantum of power within an organization that can be distributed in differing ways; and it does not account for the fact, now rather well documented, that power is highly variable. The addition of power at one level of the hierarchy does not at all mean the automatic withdrawal of power at another.73 Another problem with this conceptualization of power is that the ends of the continuum cannot represent real situa tions; total consolidation of power, represented by one end of a continuum, is impossible. If this were to occur, "...the concept of organization is denied."74 And at the opposite end of the continuum it is equally difficult to conceive of pure decentralization. "The line between 'pure' decentraliza tion within an organization and actual separation from the 75 organization is very unclear." One end of the continuum, centralization, would imply a total lack of differentiation within the system and therefore deny the concept of organization; the other extreme, decen tralization, implies total independence and consequently the lack of interrelationship and integration — in short, the denial of the concept of a system. Also, the possibility of non-hierarchical systems is denied. Sherwood argues that this is precisely what has occurred in the theory of organization. This denial has been so complete that there is simply no useful terminology to describe and explain the behavior of non-hierarchical systems. 123 This has, in turn, led to a lack of appreciation of the extent, importance and frequency of non-hierarchical acts of coordination within organizations. According to Sherwood's formulation of this problem, the meaning of decentralization should be restricted to intra organization power relationships and the term devolution should be used to describe an inter-organization pattern of power relationships.^ There is a subtle but highly important twist in logic involved in these two definitions; upon closer examination, it becomes clear these two definitions are not related to the same class of behaviors. The first definition has to do with the expression of power within a single organization. Through out the discussion the focus has continually been on relation ships within a system. This definition shifts the discussion to the consideration of one part not necessarily incorporated into a system. The second definition restricts analysis to the expression of power between two or more organizations. These are very definitely two different kinds of analysis and not an adequate way around the problem. For all practical purposes, the idea of a power continuum could still be used to analyze behavior either way. The way around this problem is to redefine the nature of power. Throughout this discussion, power consistently has been treated as a relational concept; thus, total centraliza 124 tion of power denies system differentiation and therefore denies the possibility of organization. Total decentralization implies total independence and therefore lack of relationship and the possibility of a system. Underlying this formulation is the thus far undiscussed question of what constitutes a power relationship. In order for a relationship to exist, there must be some connection between two or more entities whether they be people, organiza tions or communities. If power is defined as the extent to which one party can influence another, a power relationship can take a number of forms. One party can have the predominate influence on the relationship and the other a minimal influ ence; or, for example, both parties can have equal influence. A power relationship can be considered in terms of two distinct dimensions; (1) the degree of interdependence (mutual influence) between the parties involved in the relation ship and (2) the degree of independence between the parties. This conceptualization is presented in Figure 5. below. A centralized organization is characterized by relationships that have high interdependence but low independence. (One or more parties have a predominant amount of influence in the relationship.) A decentralized organization is characterized by a high level of independence and low level of interdepen dence . !! . 1 125 High Interdependence Low Low High Independence Figure 5.— Organizational power patterns Organizational devolution is said to exist when there Is a high level of interdependence in the relationships between the members of the organization and a high level of independence. Organizational devolution would provide for the greatest possible system adaptability and would be characteri zed by mutuality of influence. In short, a non-hierarchical power structure. Organizational Devolution as a Strategy of Organization Development For a number of reasons, organizational devolution became a strategy for change in an organizational development effort with the State Compensation Insurance Fund (SCIF). Beginning in approximately 1966, top management of SCIF started an organization development effort. This program began by hiring an outside consultant to help top management become a more effective team. The consultant did several things in an effort to achieve this goal. To help build a Centralization Devolution Anarchy (No Relationship) Decentrali zation 126 climate in which the various members would be willing to express their feelings about the organization, non-directive interviews were periodically conducted with various top managers. Gradually, as the managers grew more confident about expressing their concerns and did so, more information accumulated about how the overall organization was functioning. The role the consultant adopted was to facilitate confrontation between top managers when disagreements existed with regard to organizational policies, decisions made by the management team, or about behavior of individuals toward each other. Since it had been characteristic of the organization to invest heavily in the development of its managers (especia lly in terms of executive development programs and T-group training), it was not long before sufficient information had been accumulated, through the team building process, to cause top management to conclude that a study should be conducted to determine how the overall organization was perceived.^ The consultant conducted a series of interviews with a cross section of the organization and reported back to the top management group. It was clear to top management that many of its worst suspicions had been confirmed. Communications between various levels of the organi zation were not as effective as hoped. The field offices indicated they were having considerable difficulty getting upper levels of management to formulate new policies and 127 modify old ones quickly enough. Even though by conventional standards the decentralized structure of the organization was modern and up-to-date, it was not proving to be sufficiently flexible to meet the demands placed on the organization by both its external and internal environment. Various possible avenues of action were explored in an effort to cope with the problems of the organization. A considerable body of contemporary management literature, including the Sherwood article on "organizational devolution", was read and discussed by members of the team. Through these discussions the team became aware of the need to test a number of assumptions about how the organi zation might be designed. Heretofore, it had always been assumed that an elaborate hierarchical structure was necessary for effective operations. This proposition had never been tested; how much and what kind of structure was currently needed in the organization was not known. In an effort to test this proposition and explore the possibility of redesigning the organization, top management adopted an experimental strategy of organizational devolution. The experiment was to be conducted with one of the field offices; this role this level in the organization played was thought to be crucial to overall effectiveness of the organi zation. The experiment in devolution was to run for a year and 78 under these conditions: 128 1. The experimental office would be given a high level of autonomy. For the period of the experiment the only limits on the district office would be those that the General Manager faced. Otherwise the field office would have as much autonomy as the General Manager; any thing he could do the field office could do. 2. The district office would be provided with a lump-sum budget which could be spent, subject to the limits placed on the organization by the State of California, in whatever way the district office thought was appropriate. 3. The experimental office would be removed from its normal position in the hierarchy of the organization and report only to the General Manager. 4. All reports normally required of the field office by the larger organization would be suspended during the experiment. 5. All services — personnel, accounting, data processing, training, etc. — required of the overall organization by the field office would be obtained only if the field office was willing to contract for and pay for the services it needed out of its budget. 6. The district office would keep the few records necessary to meet the State bookkeeping re quirements . 7. Any and all organizational records were to be made available to the district office — re gardless of its character — if the district office asked for it. In return for these conditions, the district office was asked to prepare a program statement which: 1. Explained why the SCIF had originally been established. 2. Explained the current need for the organization. 129 3. Presented the broad goals the district office intended to pursue during the course of the experiment. 4. Stated the various, short-term objectives the organization intended to achieve. 5. Spelled out alternate paths to the objectives. 6. Indicated which paths to the objectives would be pursued during the experiment. In September, 1967, the experiment was begun; in the late Spring of 1969, top management of SCIF decided that these conditions should be spread to all of the 24 field offices and the task of organizational re-design was underway. NOTES CHAPTER III 1. Warren Bennis, Changing Organizations (New York: McGraw-Hill Book Company, 1966), p. 126. 2. W. WArner Burke,"Management Development and Organization Development", Journal of Applied Behavioral Science, VII (September/October, 1971),569-579. 3. Ibid., 571-573. 4. Ibid., 575. 5. Stanley M. Herman, "What is This Thing Called 'Organization Development'?" (unpublished paper, TRW Systems, Redondo Beach, Calif., approx. 1970). (Mimeographed.) 6. Robert Blake and Jane Mouton, The Managerial Grid KHbuston, Texas: Gulf Publishing Company, 1964), p. 261. 7. Warren Bennis, Organization Development (Reading, Mass.; Addison-Wesley Publishing Company, 1969), p. 2 8. Richard Beckhard, Organization Development: Strategies and Models (Reading, Mass.: Addison-Wesley Publishing Company, 1969), p. 9. 9. Ibid. 10. Cf. Chap. ii, especially the material about the effects of environmental complexity. 11. Amitai Etzioni, The Active Society (New York: Free Press, 1968), p. 66. 12. John T. Dorsey, Jr., "An Information-Energy Model," in Papers in Comparative Administration, ed. by Ferrel Heady and Sybil Stokes (Ann Arbor, Mich.: University of Michigan Press, 1962), p. 43. 13. E. L. Trist, "On Socio-Technical Systems," in The Planning of Change, ed. by Warren Bennis, Kenneth Benne and Robert Chin (2nd ed.; New York: Holt, Rinehart and Winston, 1969) , pp. 269-281. 110 131 14. Daniel Katz, "Group Processes and Social Inte gration," Journal of Social Issues, XXIII (June, 1967), 3-22. 15. Ibid., pp. 7-10. 16. Cf., Robert Blake, Herbert Shepard and Jane Mouton. Managing Intergroup Conflict in Industry. Chapter iv. (Houston, Texas: Gulf Publishing Company, 1964). 17. Ibid. 18. New theoretical developments in research methods raise serious doubts about whether the extant behavioral methods can produce reliable knowledge. Cf. Argyris, Intervention Theory. 19. Robert Blake and Jane Mouton, Building a Dynamic Corporation Through Grid Organization Development (Reading, Mass.: Addison-Wesley Publishing Company, 1969) , p. 16. 20. Robert Blake and Jane Mouton, Corporate Excellence through Grid Organization Development (Houston, Texas: Gulf Publishing Company, 1968). 21. The grid theory and research literature from which it was developed can be found in Blake and Mouton, Managerial Grid. 22. The theory behind this phase is derived from the work of the social psychologist Muzafer Sherif, who was the first person to systematically study conditions of intergroup conflict and cooperation. The book which develops the indus trial application of this theory, Blake, Shepard and Mouton, Managing Intergroup Conflict, is dedicated to Sherif. 23. Robert Blake, et. al.. "Breakthrough in Organi zation Development," Harvard Business Review, XL (November/ December, 1964) , 133-155; Robert Blake, "A Second Breakthrough in Organization Development," California Management Review, XI (Winter, 1968), 73-78. 24. Larry Greiner, "Evaluating a Pilot Program using Management Grid for Organization Development in the U.S. Forest Service (unpublished pamphlet, U.S. Forest Service, n.d.). 132 25. Larry E. Greiner, et. al., "The Simple Complexity of Organizational Climate in a Governmental Agency," in Organizational Climate, ed. by Renato Taguri and George Litwin (Cambridge, Mass.: Harvard University Press, 1968), pp. 195- 224. 26. Blake, "Breakthrough in Organization Development." 27. Ibid. 28. Ibid. 29. A "One-Shot" case study: Cf. Julian Stanley and Donald Campbell, Experimental and Quasi-Experimental Designs (Chicago: Rand McNally & Company, 1966). 30. Blake, "Breakthrough in Organization Development." 31. Greiner, "Evaluating a Pilot Program," p. i. 32. Ibid., p. 6b. 33. Edward C. Tolman and Egon Brunswik, "The Organism and the Causal Texture of the Environment," Psychological Review, XLII (January, 1935) , 43-77. 34. Greiner, "Evaluating a Pilot Program," p. 19b. The preceding material was related to learning goals of the Phase I program, not to behavior in the organization. Conse quently, this material will not be reported. 35. Ibid., p. 20b. 36. Ibid., p. 21b. 37. Ibid., p. 22b. 38. Ibid., p. 24a. 39. Ibid. 40. Ibid., 24b, 25b, 26b. 41. Greiner, "Simple Complexity • C M Ibid., pp. 201-203 43. Ibid., p. 204. 198-199, 133 44. Ibid., p. 196. 45. Ibid., p. 199. 46. Ibid. , p. 196. 47. Kenneth Benne, "History of the T-Group in the Laboratory Setting," in T-Group Theory and the Laboratory Method, ed. by Leland Bradford, et. al. (New York: John Wiley & Sons, Inc., 1964), p. 81. 48. Carl Rogers, On Encounter Groups (New York: Har per & Row, Publishers, 197fi). 49. Bennis, Changing Organizations, p. 138. 50. Ibid., p. 120. 51. Cf. C. M. Hampden-Turner, "An Existential Learning Theory and the Integration of T-Group Research," Journal of Applied Behavioral Science, II (October/November/ December, 1966), p. 367-386. 52. E. A. Fleishman, E. T. Harris and E. H. Burtt, Leadership and Supervision in Industry (Columbus, Ohio: Ohio State University, 1955). 53. R. Lippett, J. Watson and B. Westley, The Dynamics of Planned Change (New York: Harcourt, Brace & World, Inc., 1961). 54. The following article gives a remarkably lucid description of team building as it was experienced by a manager and his subordinates that worked for the U.S. State Department. Cf. William J. Crockett, "Team Building — One Approach to Organization Development," Journal of Applied Behavioral Science, VI (July/August/September, 1970), 291-306. 55. Lippett, Planned Change; Sheldon Davis, "An Organic Problem-Solving Method of Organizational Change," Journal of Applied Behavioral Science, III (January, 1967) 3-21. 56. Chris Argyris, Organization Development (New Haven, Conn.: Yale University Press, 1960). Cf. Bennis, Changing Organizations, pp. 127-129 for a summary of the work. 57. Edgar Schein, Process Consultation (Reading, Mass.: Addison-Wesley Publishing Company, 196^), p. 7. 134 58. Argyris, Intervention Theory/ pp. 16-31. 59. Ibid., p. 17. 60. The material which follows was adapted from Jack Fordyce and Raymond Weil, Managing With People (Reading, Mass.: Addison-Wesley Publishing Company, 1971), pp. 124-130. 61. Richard Beckhard and Dale Lake, "Short and Long- Range Effects of a Team Development Effort," [unpublished article in 1968 that was revised and published in Social Intervention: A Behavioral Science Analysis, ed. by Harvey A. Hornstein, et. al. (New York: Free Press, 1971)]. 62. Ibid., p. 4. 63. Ibid., p. 6. 64. Dale E. Zand, et. al., "The Impact of An Organi zation Development Program on Perceptions of Interpersonal, Group, and Organizational Functioning," Journal of Applied Behavioral Science, V (July/August/September, 1969),393-411. 65. Ibid., 400-409. 66. Stanley E. Seashore and David G. Bowers, "Dura bility of Organizational Change," American Psychologist, XXV (April 1970), 227-233; Alfred J. Marrow, David G. Bowers and Stanley E. Seashore, Management by Participation (New York: Harper & Row, Publishers, 1967). 67. Alfred Marrow, "Planning the Changes," in Marrow, p. 66. 68. David G. Bowers and Stanley E. Seashore, "Employee Attitudes, Motivations and Satisfactions," in Marrow, pp. 185-201. 69. Ibid., p. 201. 70. David G. Bowers and Stanley E. Seashore, "Im plications for Managing Change", in Marrow, p. 230. 71. Frank Sherwood, "Devolution as a Problem of Organization Theory" (unpublished paper, University of Southern California School of Public Administration, n.d. ((Mimeo graphed) ) . 72. Ibid., pp. 7-10. 135 73. Ibid., p. 7. 74. Ibid., p. 8. 75. Ibid. 76. Ibid., p. 12. 77. This intervention was described earlier in this Chapter; it can be found in the discussion on Emergent/Situa tional Change strategies. 78. Later the experiment was extended for six months. Cf. Chapter iv. CHAPTER IV A PARTICIPANT OBSERVER'S VIEW OF THE EXPERIMENT Somewhere near the first of February, 1967, I became in volved with the State Compensation Insurance Fund's (SCIF) plans to conduct an experiment with one of their field offices. Ini tial contact with the organization was made through the organi zation development (OD) consultant who had been working inten sively with the Fund for several years. We had very tentative and casual discussions about the possibility of documenting the effects of an experiment in organization devolution; these discussions continued on a sporadic basis for about one month. I did not become involved in a serious way until top management of SCIF decided to commit themselves to the experiment. Once this preliminary decision to undertake the experi ment was made, the field office selected by top management as the potential experimental organization was contacted. Actu ally, preliminary explorations between the District Manager of the office and the OD consultant went forward as early as March, 1967, prior to top management's decision. The discussion between the consultant and the District Manager was tentative and speculative. The District Manager favored the proposed experiment and explorations began in earnest very shortly. One of the conditions of the experiment was that the 136 137 District Office would develop a proposed program statement. Goals and activities to include in the statement were discussed, beginning at the level of supervisors in the District Office. Early Explorations in the Field Office Each week an afternoon was set aside for supervisory personnel— the two sales supervisors, the safety supervisor, the auditing supervisor and the clerical supervisor— to discuss the functioning of the District Office, its problems, and any thing else about the organization these people needed to share. At one such meeting the District Manager asked the supervisors what their subordinates would change if their Office could do whatever it wanted to for one year. After several hours of general discussion of this proposition, he instructed supervisors to ask their men this question. Opinions gathered from this survey would be presented at the next meeting of supervisory staff. It should be noted that the District Manager did not tell the supervisors that their office was virtually certain to be the experimental office.* Perhaps because of this ambiguity, or simply because of personal style of the supervisors, the next supervisory meeting brought forth a list of complaints in response to the supervisors' queries. The District Manager was dismayed; he had anticipated proposals for programs, not a tally of malfunctions in the State Fund. Few complaints involved personnel problems or conflicts within the District Office. Nearly all were associated with objects: office equipment, technical material, sales aids, automobiles. A few indirect references to the idiosyncratic practices of the supervisors or the District Manager appeared. The tone of the reports by supervisors at the meeting was skepticism. Almost all employees doubted that the larger or ganization was serious about enabling change to occur in the areas of traditional concern, i.e., salaries, their circum scribed opportunities for promotions, automobiles, or the divi dend scale (which would make SCIF more competitive selling to small accounts). The auditing department clearly believed its status to be unjustifiably low in the eyes of District Office staff; the sales units and the safety unit wanted to make sure that other parts of the organization could not unduly influence the operations of their units. General discussions of problems that could be solved by autonomy for the office continued for several weeks. Material from these discussions was summarized in a 2 preliminary statement and taken to San Francisco m July. There the District Manager of the Experimental Office met with top management of SCIF in the executive conference room, the Blue Room. The initial meeting was closed to the District Manager (he was asked to sit in the hall for two hours while top management decided to give its imprimatur to the experiment. This meeting revealed considerable conflict; several Managers below the level of the General Manager were quite uneasy with 139 the idea of the experiment and pursued a number of issues, each of which could have thwarted the decision to proceed. After some heated exchanges, a not altogether wholehearted consensus decided to launch the experiment if the experimental office would develop a program acceptable to a special review committee organized by top management. This meeting opened a period of contradictory require ments for all of the parties concerned with the experiment. To me, one statement effectively summarizes the nature of the agreement which was reached at the meeting: Begin, and do all you can, but remember you have not begun. The operational meaning of such a paradoxical imperative would have to be de termined by all of the interdependent parties: top management, the District Manager, and the staff of the experimental office, in the course of events before the experiment began. Following this meeting, another conclave was organized in the general conference room. This one included all top managers, the managers of the various staff services, and the General Manager. The General Manager and Neely Gardner ex plained the design of the experiment; next, the District Man ager set forth the proposed program of the experimental office. Because the experiment would involve an exploration by experi mental office staff of a number of top management, the General Manager asked everybody to cooperate fully with the experimental office. He also urged concentrated effort in the next few weeks; he wanted the experiment to be launched formally as soon 140 as possible and understood that this would require complicated rapid exploration in a number of subjects. In general, the reaction of those present at the meet ing was guarded. A few seemed enthusiastic; most were noncom- mital. Many conferees probably were noncommital because the tentative statement of the program of the District Manager was a highly charged, emotional document that clearly implied dis satisfaction with much of the work of the managers around the table. The program statement extravagantly promised improvement in organization performance. Here are three predicted results of the experiment:^ 1. Write 43 percent more new business during the pilot year than in the calendar year 1966. 2. Achieve a 15 percent net premium gain, using as as base period the calendar year 1966. 3. Create and maintain a pool of people who are char acterized by: self-awareness, growth, development, stress resistance, resilience, independence, empathy, capable of love, having perception free of distortion, and feeling-tone based interpersonal relationships. All of these attributes are encom passed by the term "self-actualization." The traditional organizational arrangements were obviously be lieved to obstruct District Office performance in extraordinary ways. It is unlikely that this implication escaped top man agers who were present at the "kick-off." After the meeting, a series of personal interviews began between the top staff members in research and systems, personnel, finance, underwriting, insurance services, claims and the District Manager. The interviews were based on the 141 premise that these department heads would spell out the limita tions they felt were placed on the General Manager of the SCIF organization. Since these would also be the boundaries of what would be possible for the experimental office manager, the interviews elicited much behavior. This encounter involved a complete role reversal for everyone present. Normally/ the District Manager was, in ef fect, told by top staff people what his limits were and what criteria would be used to evaluate the performance of his office. In general, few of the criteria were developed collabor-i atively. As in most organizations, the evaluative measures were handed down by fiat in SCIF. In these interviews, people who formerly were responsible for developing and implementing criteria for evaluation of performance were now asked to share their premises with the District Manager and, in addition, to tell him exactly how much power and maneuverability the most powerful officer in the organization could muster. No small part of the interviews explored the basis of existing evaluative criteria. When the District Manager questioned these premises, the gap between what was actually occurring in the field and what was presumed to be occurring was enormous. All traditional "hard" criteria employed to aware promotions and other signs of favor became "soft." One consequence of these interviews, besides develop ing the District Manager's consciousness of the limitations 142 surrounding the General Manager, was to confront top staffers with the fact that the traditional manes of evaluation were hardly precise, unambiguous, standards. As another consequence, the District Manager learned that much anxiety of technical service personnel about the experiment was caused by the proposed experimental office being an "unrepresentative" office; it was not a dispersed office like most SCIF offices. Any positive or negative results of the experiment might not hold for the dispersed offices, they suspected. Following this first day of meetings, the District Man ager and I traveled often to San Francisco to continue assess ing the limitations placed on the General Manager, the kind of evaluative criteria that might be used by the experimental of fice to assess its own performances, the form and quantity of statistical data available but not generally communicated to lower levels in the organization and finally, the possibilities for new programs foreseen by people in the Home Office who had the reputation of being innovators. While these visits continued, a series of special meet ings was launched in the District Office. The traditional supervisors' meeting was abolished and replaced by a meeting of the entire staff. These fall meetings were intended to broaden opportunities for communication among everyone in the District Office; to discuss future directions for the office, and to work out specific agenda. 'i 143 Meetings continued to reveal deep skepticism that top management of SCIF was serious about granting autonomy. Past incidents were resurrected as evidence that management was not sincere in proposing such extraordinary autonomy. The experiment was widely regarded as a "game" played by top management with the District Office for purposes unknown. This skepticism lingered until several weeks later when the experiment formally began; then, it slowly subsided. Since such meetings were completely novel to most people in the office, they generally were chaotic, unfocused, and ineffective. Whenever the group began to crystallize a position on some problem, great cries of alarm went up from some persons. Interestingly, those who objected were also the people who seemed most afraid of the general meetings. These two or three persons were unhappy with the concept of solving organizational problems through general meetings. "By God, it is the supervisor's job to decide what I do and nothing should happen to change that!" When other staff members were prepared to commit themselves to a course of action after discussing possible solutions at length, the standard response of these two or three was to raise doubts. They tried to show that no certainty about outcomes existed; they questioned the possi bility that anyone besides the formal managers could make decisions so long as the formal hierarchy was still evaluating the staff (and were thus "clearly" the persons to take 144 positions of authority). Their objections, again and again, dissipated trust among staff members present, forced pro crastination or ridiculed the idea that the substance of office work should be approached by anyone other than the supervisors or the District Manager. Characteristically, these persons would ask many probing questions that forced other participants to take a stand; none of them ever directly revealed where he or she stood. Nevertheless, their positions were often implicitly unmistakable because of the questions they asked. In summary, the general meetings were ineffective devices for problem-solving or planning. The two or three persons most afraid repeatedly prevented the meeting from reaching a group decision. These persons were also the most vocal opponents of the general meetings and most critical of the ineffectual decision-making that occurred there. A Bolt from the Blue The District Manager and I were interviewing one of the Home Office staff when we received a written directive to appear in the Blue Room at 11 a.m. for a special meeting. Neither of us knew what the unspecified agenda of the meeting might be, but we speculated that a reaction to the inter viewing had occurred or that a problem between the District Manager and the Regional Manager had been pushed upward in the organization. 145 The meeting, called in the absence of the General Manager, revealed confusion over whether the District Manager was overstepping his authority by traveling to San Francisco without the approval of his Regional Manager. The Regional Manager was on hand, as were the individuals who had partici pated in the meetings which had formally decided to undertake the experiment. The only top level person not present at this meeting was the General Manager who was out of town. Managers declared that the District Manager was mis reading his role; "at this point in time, all that he was supposed to do was gather information about what the District Office might do if its program was approved," said one. The Managers felt the District Manager was "confused" about what he was supposed to be doing; he was supposed to be working only on the design of the experiment, while imple mentation would occur later. The District Manager was told that if he wanted to travel to San Francisco, he should request the approval of his Regional Manager. The District Manager should not undertake or allow any activity in the District Office that was not normal operating procedure before the program was formally approved. Further, he was warned that it was incorrect to interpret the General Manager's speech at the large "kick-off" meeting to mean that the District Office or the District Manager could initiate changes in the organization before formal approval of the program statement. The fact that the experimental 146 program was to run from July 1, 1967, to July 1, 1968, and that July was almost over was immaterial. Despite the tone of this meeting, the District Manager made trips to San Francisco, and the program statement was compiled and worked out with the special program review committee. The General Manager approved it informally and at last the experiment received formal approval at a special meeting held in the Experimental Office on September 8, 1967. This special meeting, attended by the General Manager of the State Fund, was designed to authorize formally and legitimize the experimental program. The meeting was a memorable half-day for District Office staff. Launching the Experiment 4 Following the final approval meeting at the Experi mental Office (EO) weekly and often biweekly general meetings continued. In this period, a number of new subcommittees began to work closely with the San Francisco office in germinating specific programs for action from the early program statement of the District Office. The District Manager's Strategy: A Leap into the Future Meanwhile, the District Manager took certain actions and modified many of his ideas about directly converting his organization from a relatively hierarchical organization to 147 a multi-hierarchical system characterized by widely shared power and expertise. One of the urgent problems he now saw was that two of the five supervisors were not regarded as effective managers. They were not self-starters who would take responsibility for a problem and persist until it was solved; instead, they would bring problems to the District Manager for solution. These two supervisors also believed that subordinates must be told how, when, and what to do. They felt unless they built a highly directive structure and periodically evaluated the achievements of their subordinates in details, the performance of the organization would fall off drastically. With this personnel problem in mind, the District Manager (hereafter called Manager Early) decided not to tell the supervisors at the outset (see above, p.137) that their District Office was almost certain to be the experimental site. Further reasons urged him to pursue this discreet opening strategy. Not the least of these was that the nature of the supervisory structure of the EO was unique in SCIF because of the legacy of the ex-District Manager, hereafter called Manager Goode. Manager Goode had been rather paternalistic, allow ing individual staff members to bypass supervisors and come directly to him to resolve organizational problems. He would frequently make a decision, later informing the employee's supervisor that such and such would be done 148 regarding the employee. When Goode entered the office each morning he generally made a point of greeting the entire staff. When District Manager Early moved into the organiza tion, one of the first things he heard from the supervisors was that their jobs had no meaning. They felt continually bypassed by the employees and were unsure of any genuine responsibility for the way the office functioned. Conse quently, the new District Manager chose generally to avoid involvement with the office staff other than the supervisors. Weekly meetings with the supervisors were begun to air their problems. Manager Early firmly discouraged any attempt to bypass the supervisors. Gradually, according to the District Manager, supervisors began to feel that they did have responsibilities in the organization, and their jobs became meaningful. Many employees felt that Early was cold and indifferent because of the way he had handled the problem of supervision. One strategy chosen by the District Manager to place the supervisors in positions of responsibility was deliberately to stay away from the office. Manager Early spent most of 1966 and early 1967 writing a training manual to be used in all District Offices. Early tried to stay out of the office during this period while making clear to the supervisors that he would support any decisions they t 149 made in his absence. The result of this approach was, accord ing to the District Manager, to revitalize the supervisors. The District Manager was proud of his conscious attempt to "rehabilitate" persons who were rejected by other District Offices or subsystems of SCIF. A number of the employees did have lackluster records in other parts of the SCIF and were doing well in Early's District Office. Another source of pride for the District Manager was that his was the first District Office to have a woman as supervisor of anything besides the clerical staff. She was not appointed formally to that position but was instead, "acting supervisor." Working in this position for almost one year, she had achieved excellent results. Other employees murmured some discontent — although low-keyed — about a woman holding a supervisory position. As this supervisor had demonstrated that she could handle the job, every effort was made to insure that she would be prepared for the promotional examina tions and receive full supervisory rank. Manager Early was concerned about the risk this strategy involved. He felt that SCIF professionals outside the District Office who would participate in the promotional evaluation were likely to be biased against the candidate because of her sex. Finally, since Early felt that he had been initially successful in promoting risk-taking behavior and responsibility on the supervisors' part, he would attempt to minimize his own role until the organization accepted him as a peer and would not remain dependent upon him. The District Manager believed that any highly effective organization could tolerate much ambiguity and commit itself to action as a consequence of dialogue, without codified programs or goals. To the extent a commitment in writing was necessary, Early thought, the organization was not functioning well.^ Changes, Time and Revival After the special meeting in the District Office which formally launched the experiment, general meetings continued. At least two important events occurred during these meetings. Before the first one, the functional units (sales, safety and auditing), which except for the Auditing Supervisor, were all male, had opposed including the all female clerical staff. Since the ground rules of the meetings were that each participant would have equal power in decision-making, it was impossible for the District Manager or the Auditing Supervisor to overrule this argument without violating the rules to which they had agreed. After several heated discussions, during which several of the men threatened not to attend the meeting if women were admitted, it was finally decided temporarily to exclude the clerical staff. After approximately three general meetings this decision was reversed. The women of the clerical staff were 151 admitted to meetings and accepted, with more than a few reservations, as equals. Regardless of the inefficient, ineffective impression these general meetings made upon an outsider, they were very significant to the organization. Foremost among their effects was that the meetings defined the project as an attempt to build a fully participative organizational system. The impor tance of this identification of participative meetings with more open organization cannot be underestimated. The success of these meetings was, as before, generally low. Manager Early was still distrusted by the two or three most afraid persons in the office and frequently fielded hos tile remarks from these persons, who felt that Early was acting irresponsibly by not providing the one best solution to office problems. In addition, two supervisors were unsure of the re quirements of their jobs, given the new participative approach. A Major Unexpected Change The next major problem arose on November 7, when I received a phone call from the District Manager. He reported that he had been asked to resign or face formal dismissal. Immediately after this conservation, I received a confirming call from one of the top managers in SCIF.** Within two weeks the District Manager resigned. A complete audit, directed by one of the top managers from the Fund's central headquarters, of the experimental 152 office was conducted during these two weeks.^ During the audit, the old supervisory role structure, which had been in abey ance for several months, was reestablished. Top management's representative personally kept the office under lock and key, padlocking the doors each night and removing the locks each morning. This was a symbolic event which did not go unremarked by the office staff. The experiment was extended six months in order to enhance the possibility of affecting it. The person appointed to become the new manager of the experimental office had previously been the training director of the entire SCIF organization; before this assignment he had managed the district office that was the experiment's aware control group. The Task of Rebuilding The new District Manager, Mr. Beard, faced a bizarre set of circumstances. The supervisory structure of his previous organization resembled the pre-experimental structure in his new office. Beard regarded this pre-experi- mental structure favorably and, at first, thought he would attempt to install a similar system in the experimental office. Rather than move immediately, he decided to discuss matters with the supervisors and to support the continuation of the decision-making processes that were under way prior to Early's resignation. 153 One thing rapidly became apparent to him — in addition to the ineffectiveness of the general meeting format. The authority structure with which Beard was familiar had, in fact, reduced the possibility of obtaining a wide range of information about the way his organization was functioning. Because of his tolerance for ambiguity and skills in inter personal problem-solving, Manager Beard was able effectively to integrate himself into the organization and to foster a climate of trust. He was able to do so despite the fact that the scared staff members continued to perform in general meetings as they had since the experiment began. Immediately before Early's departure the staff of the EO had decided to conduct a team-building program with the help of an outside consultant. The session was cancelled because of the sudden change of leadership, but within a few weeks of Manager Beard's arrival the team-building program was revived. Virtually the entire staff voluntarily traveled to an off-site retreat for this three-day exercise in organiza tion development. Dr. E.J. Jones, an experienced trainer and National Training Laboratories Fellow, led the program, which played a very important role in building trust and clarifying many interpersonal conflicts in the organization, both longstanding and recent. In general, the staff approved this team-building effort as the most rewarding of their staff meetings. 154 The next significant experimental event was the establishment of the PERT Committee, a special committee formed through a rather strange but understandable process. The District Manager casually mentioned to one of the staff O that PERT techniques might help the organization to plan toward achieving its goals, as well as to select useful actions regarding their forthcoming move to a new location. The staff member understood this suggestion as an indirect command to bring in PERT. Consequently, he passed Beard's suggestion to several of his friends on the staff and they all began reading about PERT — unknown to the District Manager. After ten days or so, one of these persons told the District Manager that he and several others wanted to see whether PERT could help them. The District Manager responded by suggesting that they form a PERT Committee, adding one other staff member to the nucleus of their group. At approximately the same time the District Manager cancelled the large general meetings. Several weeks later it became clear to the Manager that this move was generally regarded as the end of the experimental project. These events were interpreted in this way because, in the minds of most staffers, the experiment had been defined as one in fully participative management. Ending the large general meetings that had been identified with such management, and establishing the PERT group, meant, therefore, a return to oligarchy — the traditional 155 organization. This perception was reinforced by the PERT Group, which deliberated behind closed doors for several weeks. Even though the PERT Committee was composed of persons who were highly esteemed and who made attempts to obtain information and suggestions from other members of the organization, these changes brought about an "end of the project" feeling. The committee labored to base its considerations on precisely those problems and issues which had been discussed in the earlier general meetings. After approximately four weeks, a general plan emerged and was submitted to the entire office staff for approval. After much discussion the plan was approved, and special committees were formed to fulfill it. At this point the PERT Committee was disbanded, but one of its members was selected to serve as special coordinator of activities it triggered. At first, enthusiasm about implementing the PERT program was low, but gradually it swelled. This occurred in part because it became clear to the staff that many of the goals of the program were those that the organization had earlier decided to pursue. In part, support increased simply because the PERT Committee had done its work well. I do not mean to suggest that implementation was smooth or faultless. The aloof and uncooperative attitudes 156 of many staff members in the Fund's headquarters persisted, though they have not been mentioned previously. Central staff members who were initially fascinated by the experiment (always a minority) remained supportive and helpful throughout the one and one-half years of the experiment. Staff members who were initially dismayed by the experiment generally remained uncooperative and subtly resisted providing informa tion and services to the experimental office. Lack of coordination between central staff and experi mental office activities led to more than one instance of duplication: both systems gathered information about the same problem and then reached completely different conclusions. Consider the case of territorial alignment. Unknown to each other, both the central staff and the experimental office were making a study of SCIF's market conditions and potential. This began in the central staff before the experiment; it was begun by the field office when the experiment was about half over. Both organizations reached their conclusions at about the same time, but since the central staff had begun its market analysis earlier and its decision affected the territorial alignments of all 24 field.offices, the decision of the central staff was imple mented despite considerable protest from the experimental office. The decision decreased morale. Later, too, it became clear that the decision of the field office was better than the central staff's. 157 Such events, which occurred several times, evoked almost continual questioning in the experimental office of the authenticity of the experimental grant of autonomy. Despite these problems, a number of interesting changes had occurred in the office by the end of the experi ment: The Development of a Special Public Agencies Unit This unit was formed expressly to improve the quality of SCIF's service to the various public agencies in its clientele. The unit responded directly to the organization's felt need (discussed earlier) to remain competitive by improving this aspect of its activity. Born of discussions held in the district office, the special unit was not suggested or imposed by the Fund's central office. Matrix Organizational Teams Two general problems emerged early in the general meetings. In some ways, the most important was that the functional units (sales, safety, auditing and clerical) had been elaborated into counterproductive, from the point of view of the organization as a whole, stereotypes of each other. As highly cohesive groups, they resisted external suggestions for change and were often engaged in thinly veiled status rivalries. 158 At first the major complaints of clients about the uncomfortably high turnover of field personnel handling their accounts was not regarded as extremely important. Later the complaint loomed larger and larger. From the earliest discussions of this issue, it was clear that the problem was partly created by the organization's practice of rotating personnel from function to function in order to train generalists and to minimize misunderstanding among functional units. No one wanted to stop the practice. The problem was how to resolve the clients' problem without giving up job rotation. After exploring both problems, the organization decided to build inter-functional, or interdisciplinary matrix teams to serve its clients. Each of the several teams included at least one representative from sales, safety, claims, auditing and the clerical staff. Each team chose one person to act as general coordinator of the activities of the matrix unit. Employment of this structure had a number of desirable effects. It helped clients to feel that they received better, more systematic service for their money. For the staff it promoted interdisciplinary interdependence and cooperation; the job rotation program could continue without disrupting client relationships; rotation now took place within the teams and was explained to and accepted by the clients. Since each team was given a specific territory 159 and held responsible for it, matrix teams proved effective devices for integrating the various functional disciplines into a coherent, interrelated program package. Matrix teams did not, however, supercede the old disciplines. Once a week the members of each discipline (sales, safety, claims, auditing and clerical) held a general meeting and training session to discuss problems and new developments in their field. Absorption of a New Organizational Function An additional achievement of the experimental unit was the relatively smooth absorption of a new, fairly large functional unit (claims). Claims joined the matrix teams without major problems or disruption of client services. Goals and Commitment An additional consequence of the experiment was the emergence of clear, well-defined goals and programs. Most importantly, an effective system was built to assess progress toward goals. The members of the experimental organization felt, by contrast to their colleagues at other SCIF field offices at this time, reasonably strong commitment to the goals of the organization. Personnel Turnover The turnover rate in the experimental office had been about 30 percent per year during the course of the experiment. 160 I think it is interesting that all persons leaving the organization left because they felt they had discovered what they really wanted to do with their lives and that they could not do it within the SCIF organization. In addition to such cases, several promotions to other jobs in SCIF sent staff members to other locations and to jobs that were more closely akin to their interests. Unanticipated Consequences of Crisis One of the spin-offs of the leadership crisis of Early's departure in November was a sudden and completely new level of support for the concept underlying the experi ment: that SCIF field offices could operate effectively without the traditional organizational hierarchy. Within a short time, the nay-sayers had become the most fervent yea-sayers. As this transformation occurred within the upper levels of SCIF, a decision was made to move the entire organization into a new multi-hierarchical structure. By late Fall, 1969, all District Offices in SCIF had the same operational freedom as the experimental District Office. The pioneer experimental office asked the General Manager of SCIF to share the experience of formally closing the experimental program. Returning the office to the "normal" SCIF organization was accomplished on December 19, 1968, one and one-half years after the program began. NOTES CHAPTER IV 1. The reasons why he did not will be discussed later in this Chapter. 2. Appendix I. 3. Ibid. 4. The letter from the General Manager announcing the experiment to all parts of SCIF is found in Appendix IV. 5. I found this argument to be persuasive also and, as a consequence, during the early pre-program explorations, sided with the District Manager and argued against Neely Gardner's proposal that the organization should develop a program budget and program statement which could serve as a device for providing objective, non-evaluative feedback on progress made during the experiment. 6. I met two Home Office Managers the next day, discussed the problems Early's resignation would create for the experimental study, and agreed to continue to document the experiment. 7. It may not have been accidental that this Manager was the one who had most vigorously opposed the selection of this field office and its District Manager as the experimental unit. 8. Program Evaluation Review Technique. CHAPTER V RESEARCH METHODS The purpose of this Chapter is to set forth the re search design, data-gathering strategy and statistical tech niques. In each of these sections, especially the second, strengths and weaknesses of the approach will be discussed in detail. Research Design In a special sense, it may be that research design is the most important part of assessing organizational experiments; this is true only because it has received such slight attention. The field abounds with what is known technically as One-Group Pretest-Posttest Design, which is inherently very weak.^" And the other very common design, 2 the One-Shot Case Study, is even weaker. From the standpoint of research theory all three aspects of research methods— design, data-gathering strategy and statistical techniques— are equally important. If one or more is poorly done, research findings are virtually meaningless. In part the problem of poor research design is simply a consequence of dilemmas which arise in field studies. To a large extent weak research designs are an unchangeable 163 consequence of the value placed upon studying existing organi zations. Almost inevitably, there are many practical problems 3 the researcher must attempt to solve as best he can. Generally, these problems are solved less satisfactorily than they would be in laboratory experiments where extraneous variables are more readily controlled and manipulated. Lack of random selection of experimental groups is, perhaps, the most notorious problem. When studying an organization it is extremely difficult to either find or create circumstances in which experimental groups can be selected at random; far more often than not, only some parts of the organization are candidates for experimentation and this population is normally so small and variable that the researcher is forced to use matching techniques rather than random selection. Matching techniques, even under the best of circumstances, are subject to a number of limitations, particularly the extent to which experimental results can 4 be generalized to larger populations. Selection was very much an issue with top management of the State Compensation Insurance Fund (SCIF). Even though there were 24 field offices, the number of organiza tions which could be considered as potential candidates was about ten. Of these, only three or four were possibilities. One was picked to be the experimental office; a second was selected to be the aware control— i.e., its members knew it was a control. An unaware control was also included. 164 Although top management had decided which organizations they wanted to involve in the experiment and how they were to be involved, it was necessary to gain the overt support and cooperation of both the experimental and aware control organi- zations. Once this had been done, five additional organiza tions were included in the data-gathering activities. Each of the five was asked to provide research information at specific times. These last five organizations were added in an effort to disguise the identity of the unaware control. To the extent this was accomplished it would be possible to compare the aware control with the unaware and determine to what degree the performance of the aware control was affected by its being used as a reference point in the study. For clarity, special symbols will be used to portray the basic parameters of the design. In the material which follows, I shall rely on the work of Julian Stanley and Donald Campbell.^ The symbols they use are: X = experimental or treatment variable 0 = observation, data gathering at a specific point in time R at the end of a row = random selection of the sample population Series of dashes (--- ) between rows = non-random selection. Any X's or 0's in a given row indicate the application of these treatments to, or observations of, the same population. Moving from left to right in a row represents movement in 165 time from a beginning to an end. Vertical alignments of these symbols, either X or 0, indicate that the events occurred at the same time in different populations. Using this notation, the research design is shown in Figure 6 below: T1 T 2 T3 T4 Experimental 0 Organization X 0 0 0 Aware 0 Control 0 0 X 0 Unaware 0 Control 0 0 X 0 T1 = Fall, 1967; T^ = Summer, 1969 T2 = Spring, 19&8; T3 " Fall, 1968 Fig. 6.— Longitudinal before and after design with aware and unaware control groups The table shows that data (0) were gathered from each of the organizations four times; the first block of data was gathered before the experiment began— which is indicated by the presence of an 0 before the X in the first row. The dashes between rows indicate that the three organizations were selected by matching techniques; finally, the X's in the last two rows reflect the fact that the experimental condition was extended to the Aware Control and Unaware Control in the interval between the last two occasions of data collection. 166 Stanley and Campbell also have developed criteria to assess the strengths and weaknesses of research designs. They have identified fifteen factors which, if not taken into account by the design, can reduce the validity of the research findings— regardless of the rigor of the methods of gathering data and the statistical analysis used. They call these factors "threats to validity" and distinguish two general types— Internal and External. The nine "Internal" factors are:** 1. History.— events, other than the experimental treatment, occurring between pre-test and post test and thus providing alternate explanations of events. 2. Maturation.— processes within the respondents or observed social units producing changes as a function of the passage of time per se, such as growth, fatigue, secular trends, etc. 3. Instability.— unreliability of measures, fluctua tions in sampling persons or components, autono mous instability of repeated or "equivalent" measures. (This is the only threat to which statistical tests of significance are relevant.) 4. Testing.— the effect of taking a test upon the scores of a second testing. The effect of publication of a social indicator upon subsequent reading of that indicator. 5. Instrumentation.— in which changes in the cali bration of a measuring instrument or changes in the observers or scores used may produce changes in the obtained measures. 6. Regression artifacts.— pseudo-shifts occurring when persons or treatment units have been selected upon the basis of their extreme scores. 7. Selection.— biases resulting from differential re- cruitment of comparison groups, producing different mean levels on the measure of effects. 167 8. Experimental mortality.— the differential loss of respondents from comparison groups. 9. Selection-maturation interaction.— selection biases resulting in differential ratesof "maturation" or autonomous change. These nine variables plus the variables Stanley and Campbell call "threats to external validity" constitute "State of the Art" thinking about research design in the social sciences. Since I am reporting the effect of an experiment in a field study or quasi-experimental situation in which it will not be possible to assume in any rigorous way that the findings can be generalized, 1 shall restrict myself to consideration of the nine threats to internal validity and show how these threats are or are not controlled by the research design. We need to take one additional step before this analy sis, that is, to indicate what kinds of strategies are relevant to controlling or reducing the impact of these threats to meaningful interpretation of research data. There are a series of strategies implicit in the work of Campbell and Stanley which we shall attempt to make explicit. According to our analysis, the nine threats to internal validity can be grouped into three areas: (1) factors affected by the presence of control groups; (2) factors affected by the manner in which measurements are taken, and (3) a factor which cuts across both areas. 168 Threats Reduced by: Conditions related to the presence or absence of control groups 1. History Addition of one or more control or nontreated groups, preferably selected at random. or Use of data collected over an extended period of time, or both strategies. If only the latter strategy is used, it is critical that the measurement process remain the same. (Cf. analysis of threat #5 below.) 2. Maturation Addition of one or more control groups, preferably selected at random. 4. Testing Addition of at least one control group which is not pre-tested, assuming both the experimental and control group are selected randomly, or addition of at least two control groups— one is pre-.and post tested, the second is post tested only. (Assuming random selection processes, differences between the first and second control group are related to testing.) 6. Regression Artifacts Avoidance of the use of groups which are extreme, either high or low, in relation to the general population as determined by some measurement device. 169 Threats Reduced by: 7. Selection Random selection of study group (s) and plus 9. Selection-Maturation Interaction examination of recruitment, selection, and turnover figures of groups which have existed before the experiment. Conditions related to the Measurement Process 3. Instability (Assuming the presence of at least one control group.) Statistical analysis and probability theory plus careful design of measurement factors. 5. Instrumentation (Assuming the presence of at least one control group.) Not modifying measurement instrument during the study and thorough analysis of comparability of "comparable'' or "equivalent" measurement methods, if these are to be utilized. General condition 8. Experimental Mortality Random selection if group(s) is (are) to be temporary plus (in the case of permanent groups) careful analysis of mortality rates in the population studies before and after the experiment. Fig. 7.— Strategies for minimizing threats to validity I j 170 Each of these strategies represents a way in which threats to internal validity can be reduced; lacking the use of the appropriate strategy, the threat to validity is given full play. Such a situation would permit the explanation of the apparent effect of any experiment by raising fundamentally unchallengeable arguments that the effects were a function of another variable rather than of the treatment or experimental situation. Thus, it is appropriate to ask what strategies for controlling threats to internal validity were present in the research design. The research design used in the study provides effective control for events which are not associated with the experi ment itself (History). All three organizations exist in a similar environment and are exposed to the same political, social, economic, and cultural processes and events. There is also no reason to believe that maturation would differ among these organizations and confound the inter pretation of research data; all three of the organizations operate within the Civil Service regulations of the State of California and all have the same personnel regulations; all three recruit their employees from the same general population and all three came into being at roughly the same time. There are, however, slight differences in clientele. The threat of Instability is controlled by the statis tical analysis; the statistical procedures will be explained in the third section of this essay. Given the kind of data 'I ' i i 171 collected, the instruments used to collect it and the analysis which will be made, this threat is also controlled by the research design. The fourth threat, Testing, is primarily a problem when many measurements are repeated in a short time. Also, it can be a problem if the population to which it is being applied knows how measurements relate to them. Since the research data were gathered at intervals of about six months, I doubt this threat had any serious result. With six-month intervals, a relatively complicated data-gathering instrument, and many intervening events, it is not likely that the respondents recalled the details of the research instrument. The next threat, Instrumentation, is a problem only when the technique of measurement is changed from one measure ment to the next; in this study, we did not change the research instrument or scoring procedures and, therefore, Instrumentation is not a problem. Regression Artifacts always must be considered when matching procedures are used in the selection of groups to be studied. But, this variable is a serious problem when groups or populations are selected on the basis of extreme scores. In such a circumstance it is quite predictable that some groups received high— or low— scores simply by chance and will get more ordinary scores next time. Thus, there may occur changes which can be interpreted wrongly as a con sequence of experimental conditions. In the SCIF study, 172 the organizations were not chosen because of extreme scores; they were selected on the basis of comparability and, while they were considered to be effective, they were not the three highest performing organizations of the 24 field offices. The variable of Selection has by now been explained through the previous discussions; that is, the groups were matched as closely as possible, they were not selected at random, and were not selected because of high scores on a particular measure. Consequently, the Selection threat can be said to be controlled in a questionable manner; clearly this situation is not what one would hope for but is nonetheless better than having no possibility of control over this threat. Experimental mortality can be a problem if it is not taken seriously. In this study, we set up special meetings to administer the questionnaire. This meant making arrange ments with the managers of the field offices to make sure that all members of the organization were present at the same time and completed the questionnaire at the same time. There was much respondent mortality during the two years of the study; however, the rate of mortality was essentially the same in all three organizations. Consequently, this threat did not apply. The last threat to internal validity, Selection- Maturation Interaction, is only partially controlled because selection is not random. As I indicated earlier, I do not consider maturation to be an issue. Since this particular 173 threat is a consequence of the interaction of biases intro duced by selection procedures and differential rates of maturation, and it cannot be said that selection processes were completely controlled, the potential effect of this threat to validity is unknown. In short, the effects of this design are: aPositive Control bControl Questionable Fig. 8.— Threats to internal validity Compared to many designs used in field studies, this one incorporates rigorous controls over the circumstances which make impossible straightforward interpretation of the data.7 Rival explanations could be given only because the threats of Selection and Selection-Maturation Interaction were not controlled. It is not possible to make these rival argu ments persuasive; change or lack of change in the research variables can be almost completely attributed to the experi ment. 174 Data-Gathering Strategy Earlier in the text (Chapter I) I discussed the focus of the study, authority structures. I gave the variables which would be measured, and their definitions. I did not give details of how each of the variables were to be measured. I shall do so in this section. This information will be provided and I shall discuss the effectiveness of the procedures. When this research began in the Summer of 1967, I decided that the most efficient way to gather data would be through a questionnaire and through acting as a participant observer in the experimental organization. (The results of the latter effort have already been reported in Chapter IV.) Having decided upon the type of research instrument, the next issue was measurement of the appropriate variables. My posture toward research was behavioral. Consequently, an important concern was: what were the most reliable and valid attitudinal measures in the literature which were theoretically applicable to the study. I read the literature systematically and identified many attitudinal scales. After considerable deliberation I chose a subset of this group. These ten research variables are described below in terms of the scaling techniques and methods of construction. All ten are designed to produce a scale score which is composed 17 5 of a subject's score on a number of individual questionnaire items. Expectations of Leaders 1. Rules Orientation.— Based on six items. Five choices are provided for each item; choices ranged from "somewhat prefer" to "strongly prefer he not." Scale score is provided by the sum of individual item responses. If number of zero responses exceeds three, no score is given. 2. Nurturance of Subordinates.— Based on six items. Same choices and same method of obtaining scale score as above. 3. Closeness of Supervision.— Five items; otherwise, same as the first variable above. 4. Universalism.— Five items otherwise same as the first. 5. Promotion-Achievement Orientation.— Five items; otherwise same as the first. Each of these scales was developed through analysis of responses of 381 subjects to a 36-item questionnaire. The questionnaire was factor-analysed and yielded the 5 variables described above. These 5 accounted for 99 percent 9 of the common factor variances. Task Characteristics 6. Job Tension.— Based on 14 items; 5 choices are provided for each item; choices range from "never" to "nearly all the time." Response alternatives were assigned coding values of from 1 to 5 respectively. Scores on 14 items summed to provide total score.10 If number of zero responses exceeds 4, no score is given. 7. Work-Related Change.— Based on 9 items; 5 choices provided for each item. The choices for the first item ranged from "is always the same" to "change a great deal." Choices for the second through ninth item ranged from "I strongly disagree" to "I strongly agree." Responses 176 coded serially from 1 to 5 with high item scores indicating favorable change attitudes. If number of zero responses exceeds 3, no score is given. Corrected odd-even reli ability coefficient was .79. Items meet the criteria for a Likert scale.1! 8. Participation in Decisions.--Based on 7 items; 5 choices provided for each item. Individual item responses summed to obtain scale score.12 If number of zero responses exceeds 3, no score is given. Self at Work 9. Self-Acceptance.— Based on 33 items; 5 choices provided for each Item;choices ranged from "Not at all true of myself" to "True of myself." If number of zero responses exceeds 7 no score is given. Developed using Likert procedure. Split-half reliabilities of .894 or greater in 5 applications. 10. Aco ' “ *' ’ > ------. e 5 a p £ J J. L 1 U U 3 . All of the above scales are constructed according to contemporary scaling methodology; they contain many reversed items to reduce the possibility of the respondent's falling into a set response pattern; careful statistical analysis was used in constructing the composite scale items and in a number of instances the developers of the scale tried to determine the reliability of the items. In one instance (9 and 10), the developer also ran a series of experiments to determine the validity of the scales. science, these are exceptionally well-developed attitudinal measures. At the same time, there are a number of serious as lt-half reliabilities of .776 to .884 in According to almost any standard of behavioral !( 177 problems associated with this procedure of selecting, develop ing and administering attitudinal measures. All of these problems are associated with the relationship between the researcher and the research subjects. This would seem at first to be insignificant. Rigor ous scientific research in the behavioral sciences is con ventionally rationalized on the grounds of its objectivity. Roughly translated, the social scientist must make every effort to minimize biases in every aspect of the research process and, in particular, minimize the ways in which the subject of research can bias the measurement process. Follow ing on this rationalization is the concept that bias is re duced most effectively when the subjects' knowledge of the techniques of measurement is held to the absolute minimum. In essence, the goal is to eliminate the subjects' involve ment as thoroughly and completely as possible and thereby get reliable information. During the last several years, several authors have raised serious doubts as to the appropriateness of the now 14 classical behavioral orientation. In many ways, as Severyn Bruyn indicates, these current issues have ancient roots.^ At the same time, these issues take, on new meaning today and are especially important to a study such as this one. 178 The crux of the problem is this: the relationship between the social scientist, who is conducting research, especially empirical studies of organization development, and the research subjects can have unintended consequences for both the quality of the research data and the effort of organization development itself. And, as we shall see, this relationship is extremely important when the organization studied exists in a turbulent environment. Perhaps the most cogent formulation of the traditional behavioral approach which I have been discussing has been described by Chris Argyris in his recent book, Intervention Theory and Method.^ When he began to explore the theory, method and procedures involved in what is normally considered to be rigorous research, he discovered that most of the underlying value premises were surprisingly consistent with the values associated with bureaucratic organizations. The emphasis is upon (1) clearly defined objectives, (2) rationality, (3) simplification and specialization, and (4) the direction of control. Relevant rewards and penalties are in the hands of the researchers (the managers). Thus the subjects are dependent upon, and subordinate to, the researchers. They are given little, if any, responsibility in defining the objectives, paths and rewards. In some cases the true objective may even be withheld or distorted in order to guarantee obtaining valid data. The conditions are remarkably close to those one would find in the design and birth of a formal organization in a factory, a bank, or a govern ment bureau.17 He goes on to document a variety of ways in which the researcher subjects the members of an organization being studied to elaborate and depersonalizing controls. This, in 179 turn, has a number of unintended consequences. In summary, 18 they are : 1. Physical withdrawal which results in absenteeism and turnover. 2. Psychological withdrawal while remaining physically in the research situation. 3. Overt hostility toward the research. 4. Covert hostility as a safer adaptive mechanism. 5. Emphasis upon monetary rewards as the reason for participation. 6. Unionization of subjects. In addition, he suggests that the degree to which one or more of the above conditions is present when research is . . 19 conducted is a function of seven conditions. For our purposes, the most important of these is: "The degree to which being dependent, manipulated, and controlled is natural 20 to the lives of the subjects." This is especially pertinent to the study of organization development (OD) efforts; a central premise of OD (Cf. Chap. Ill) is the importance of creating organizational circumstances in which interdepen dency rather than dependency, self-control rather than unnecessarily restrictive external control, can be realized. These conditions are important to increasing the general level of creativity and adaptabilities of organizational members which, in turn, may increase the effectiveness of an organization in actively engaging complex internal and 21 external problems. Moreover, these conditions are ! ! 180 especially important in a highly turbulent environment (Cf. Chap. II). Based on Argyris' work, there is reason to believe the behavioral approach to measurement may not produce valid information and may have equally unintended effects upon processes in organizations. In the case of an OD study, these methods could be 22 especially ineffective ways of gathering data; and, if the OD program is taking place in an organization that is changing from either a centralized or decentralized bureau cratic structure to a non-bureaucratic structure, the research procedures may reinforce patterns of behavior contrary to what the OD program is intended to develop. Considering the circumstances of the State Compensation Insurance Fund, the OD program, and the research approach— conventional behavioral methods— there is reason to suspect that many of the issues I have been discussing are relevant to this study. My experience in gathering the data deepened my sus picion. Many people were made quite anxious by the prospect of responding to the research questionnaire; on many occasions they expressed overt hostility toward the research effort. And, finally, as will be clear in the next section of the essay, the respondent mortality rate was quite high. At the same time, the scales which were adopted for the research are, 181 at least by conventional standards, quite respectable tech niques for gathering data. The research design is rigorous and the statistical techniques are appropriate and efficient ways of assessing the data. In the interval since the data were collected these new theoretical developments, which I reported above, have raised serious issues about the methods which were employed in the study. This critique of the behavioral methodology was incorporated to ensure that the problems of this approach were clear to the reader and enable a better understanding of the research findings which will be reported in the next Chapter. Statistical Analysis After the data were gathered an analysis was made of respondent mortality. As a consequence of this effort, I discovered that very few respondents completed the question naire all four times it was administered. At the same time, the total number of respondents in each of the organizations who responded to the questionnaire during each administration did not fluctuate markedly, except when the fourth bloc of data was obtained. The total number of respondents in all three organizations— the experimental, aware control and unaware control— then increased sharply because a claims unit was added to each of the district offices. From the perspective of statistical analysis, the general instability of the research population and the small 182 number of four-time respondents constituted a serious challenge. From the perspective of organization theory, the problems seemed somewhat less significant. There is reason to believe that processes of organizational socialization quickly introduce new members to the patterns of behavior considered appropriate by the old members. This relatively enduring feature of organizations is called organizational 23 culture or organizational climate. To the extent this feature of each of the organizations was effective in introducing new members to the norms of each organization, the consequence of introducing new respondents would be diminished. Since there were two distinct groups of subjects that could be examined— the aggregate of persons who responded to the questionnaire during each administration and the sub-group who responded all four times the questionnaire was admin istered, I decided to apply statistical analyses to each of these populations. And, since the claims units which were added to each of the organizations during the summer of 1969 were not, at the time the questionnaire was administered, housed in the same physical plant as the balance of the organization and interacted very little with it, I decided to delete them from the study. This procedure left sample populations approximately equal to the populations in the earlier three administrations of the questionnaire. 24 After lengthy consultation, a statistician recom mended the above procedure and an equalization of the number 183 of respondents in each of the groups to be analyzed— the total population of respondents and the four-time respondents. It was simple to equalize the number of four-time respondents; there were nine such persons in the experimental organization and eight in each of the other two organizations. One respondent, chosen at random, was deleted from the experimental organization. Equalization was more complicated with the aggregate population. The total number of respondents in the experi mental organization during each of the four-time periods ranged from 22 to 28. The number of respondents in the aware control ranged from 17 to 24; in the unaware control, from 19 to 29. The number of respondents to be used as the equaliza tion point was 22. This was the actual number of respondents in two instances when the data was gathered. This number also minimized the amount of respondents that would have to be deleted from the study. In those instances where the number of respondents exceeded 22, the first step was to determine exactly how many had to be deleted; once this was established, all four-time respondents were withdrawn from consideration. This would maintain the integrity of the most stable of the sub populations and would ensure that the least stable would be the ones considered; presumably, this would, to a limited extent, reduce the potential affect of having new respondents during each administration. !! 184 Next, the number of persons who were involved only during the specific time period were identified. If this group held less than or equal to the number of persons to be deleted, the one-time respondents were dropped from the study. If the number of one-time respondents was greater than the number desired to be deleted, the appropriate number was selected at random and dropped from the study; the remain ing one-time respondents were combined with all persons surveyed in the organization during the time period. If the number of one-time respondents was less than the number which needed to be deleted, all one-time respondents were dropped and the rest of those to be deleted were selected at random from the population or respondents that remained— excluding the four-time respondents. Except for two instances, these procedures achieved the desired equalization. In two cases the population of respond ents fell below the desired number. A procedure which is analogous to adding additional respondents whose scores were 25 the mean of group scores was used. In one instance, this meant the addition of five "cases" and in another it meant the addition of three. The total respondent populations for each of the organizations before these procedures were used is provided in Table I below; the total populations obtained after the procedures is given in Table II. 185 Organization Fall 1967 Spring 19.68 Fall 1968 Summer 1969b Experimental 28 30 22 26 Aware Control 22 25 17 24 Unaware Control 29 25 24 19 TOTAL Claims unit excluded TABLE I POPULATIONS OF RESPONDENTS Fall Spring Fall Summer Organization 1967 1968 1968 1969 Experimental 22 22 22 22 Aware Control 22 22 17a 22 Unaware Control 22 22 22 19b aFive cases added through special procedures. Three cases added through special procedures. TABLE II RESPONDENT POPULATIONS AFTER EQUALIZATION PROCEDURES 186 The reason for using this equalization procedure was to permit a specific kind of statistical analysis— Analysis of Variance. This test was selected because it was the most efficient and generally most powerful. Analysis of Variance lends itself to a longitudinal study with more than one popula tion. By using this test, only one series of computations is necessary to test the importance of a given research variable during all four-time periods. Thus, one test could be used to determine whether or not a particular research variable was affected by the experiment. More specifically, two types of analysis of variance were used. The first of these, whose technical name is AVAR23, 2 6 was applied to the larger population of respondents. Here after, this test will simply be called "Two-Way Analysis of Variance" and the population to which it was applied will be called "Total Respondents." This particular test rests on the assumption that the organizational scores obtained during a particular time period sure independent of the scores obtained during other time periods. It is designed to com pensate for non-responses to a specific research variable, including missing cases. Because this test assumes independence between scores obtained for each of the time periods, it is not perfectly suited to the research data, the data are not completely in dependent from time period to time period, despite the problem of respondent mortality. t ! 'I 187 The only practical consequence of the assumption of independence is that the results of statistical analysis will be slightly more conservative than if independence were not assumed. The second type of analysis of variance is called "Groups-by-Trials Analysis of Variance" and will be applied to the populations I will label "Four-Time Respondents." This particular analysis is perfectly suited to the study. All of its underlying assumptions and procedures are congruent with 27 the nature of this research effort. Both of these statistical measures yield three statis tical test scores: Between Groups Variance, Within Groups Variance and an Interaction measure. The first of these is based on a comparison of each group taken over the entire four-time periods with each of the other groups over the same period. It produces a measure of the extent to which the groups were or were not similar. The second measure, Within Groups Variance, aggregates all three organizations' scores on a test variable during each specific period and assesses the extent to which the periods varied from each other. Lastly, the Interaction measure analyzes the extent to which the attitudinal scores were a product of both between and within group variance— a product of the particular organiza tions and of the time periods. To facilitate comprehension of this statistical material, I designed a special format; it will provide a graph 188 of the average scores for each organization for each time period and the supporting statistical data used to determine the between, within and interaction test scores. One such table will be used for the Total Respondents group for each of the ten research variables; a comparable set of tables will be used for the Four-Time Respondents. Results of the statistical analysis of each variable for both of the study groups will be reported together. Throughout the .05 level of significance will be used. The analysis which has just been discussed will be presented in the next Chapter. Summary This Chapter has presented the research design (Longitudinal, Before and After Design with Aware and Unaware Control Groups) and discussed its strengths and weaknesses from the viewpoint of nine different threats to validity, i.e., threats to unambiguous interpretation of the research findings. I said the design was effectively controlling for seven of the nine threats and that it offered unknown control over two of the threats— Selection and Selection-Maturation Interaction. I discussed the methodology behind the attitudinal scales and the various procedures by which they were developed. I offered criticisms of the method used in selecting, develop ing and administering these scales. I considered these 189 " criticisms especially relevant to this research study and to my experience in collecting the data. Despite the problems generated by this analysis/ I shall continue to use the assumptions with which I began, but I advise the reader to bear in mind the criticisms when the research data are presented. Finally, I described in detail the methods used to transform the research data and subject it to statistical analysis. I described the analysis to be undertaken in the next chapter, of two different populations, the "Total Number of Respondents" and the "Four-Time Respondents." I NOTES CHAPTER V 1. Donald Campbell, "Factors Relevant to the Validity of Experiments in Social Settings," Psychological Bulletin, LIV (April, 1957), pp. 297-300. 2. Ibid. 3. W. Richard Scott, "Field Methods in the Study of Organization," in Handbook of Organizations, ed. by James March (Chicago: Rand McNally & Company, 1965), pp. 261-303. 4. Julian Stanley and Donald Campbell, Experimental and Quasi-Experimental Designs for Research (Chicago: Rand McNally & Company, 1966). 5. Ibid. 6. Donald Campbell, "Reforms as Experiments," Ameri can Psychologist, XXIV (April, 1969), 411. This article contains an expansion of the list of validity factors in Stanley and Campbell. 7. Cf. Stanley and Campbell, for many examples of less powerful research designs. 8. A total of 14 different scales which produces 39 different measures was incorporated into the questionnaire and 15 non-scale control variables. Only 10 of the 39 experi mental measures are reported in the study. 9. Robert Kahn, et al., Organizational Stress (New York: John Wiley & Sons, Inc., 1964), pp. 152-153. 10. Ibid., pp. 424-425. 11. Don A. Trumbo, "Individual and Group Correlates to Attitudes Toward Work-Related Change," Journal of Applied Psychology, XLV (January, 1961) , p. 339. 190 191 12. Victor Vroom, Some Personality Determinants of the Effects of Participation (Englewood Cliffs, N. J.: Prentice- Hall, Inc., 1^60). 13. Emanuel Berger, "The Relation Between Expressed Acceptance of Self and Expressed Acceptance of Others," Journal of Abnormal and Social Psychology, XLVII (October, 1952), pp. 778-782. 14. Severyn Bruyn, The Human Perspective in Sociology (Englewood Cliffs, N. J.: Prentice-Hall, Inc., 1966); Chris Argyris, Intervention Theory and Method (Reading, Mass.: Addison-Wesley Publishing Company, 1971), esp. Chapters IV and V; Frank Friedlander, "Behavioral Research as a Transactional Process," Human Organization XXVII (Winter, 1968) , pp. 369- 379. 15. Bruyn, pp. 2-9 and 23-50. 16. Argyris, pp. 89-102, a similar point is made by Friedlander, p. 371. 17. Argyris, p. 91. 18. Argyris, p. 94. 19. Ibid. 20. Ibid. 21. Richard Beckhard, Organization Development: Strategies and Models (Reading, Mass.: Addison-Wesley Publishing Company, 1969) , pp. 31-33; Herbert Shepard, "Changing Interpersonal and Group Processes," in James March, ed., pp. 1115-1143. 22. Argyris proposes a number of ways to alter con ventional research methods to what he calls "organic" research. This approach, as yet untried in O.D. contexts, seems to be one possible way to effectively gather valid information. Cf. Argyris, pp. 100-126. 23. Cf. Renato Taguri and George Litwin (eds.) Organizational Climate (Cambridge, Mass.: Harvard University Press, 1968); additional documentation of the effect of organizational climate can be found in Alfred J. Marrow, David G. Bowers and Stanley E. Seashore, Management by Parti cipation (New York: Harper & Row, Publishers, 1967) . 24. Professor W. J. Hansche, Tulane University. I am deeply indebted to Dr. Hansche for his advice and assistance. 25. Donald J. Veldman, Fortran Programming for the Behavioral Sciences (New York: Holt, Rinehart and Winston, 1S57T7 P. 257.----- 26‘ Ibid., pp. 257-268. 27. Ibid., pp. 247-257. CHAPTER VI RESEARCH FINDINGS In this chapter, I shall analyze some effects of the experiment in organizational devolution. This analysis will be organized around the ten research variables described in Chapter I. The variables are grouped under three headings: expectations of leaders, task characteristics and the self at work. I shall use one basic format to present the research findings under each of these headings. The first part will briefly discuss the variables; the second will reintroduce the research hypotheses and summarize the reasons for their formulation; and the third will interpret the research findings. The interpretation is compiled for several reasons: the two general populations (All Respondents and Four-Time Respondents), each of which is composed of three sub-groups (an experimental, aware control and unaware control), the four blocks of data, and the statistical analysis. The statistical analysis rests on three different tests The first test, between-groups variance, yields an estimate of the probability that the three groups, taken separately and over the entire two-year period of the study, are differ ent from each other. The second test is based on a comparison 194 of all data gathered during each of the time periods; it provides an estimate of the probability that the time periods are dissimilar. The third test gives an estimate of the likelihood that the groups and the time periods are not inde pendent of each other. This is called an interaction test. If any of these measures provide a probability estimate of less than .05, I assume a significant effect exists. In other words, the groups are different from each other, the time periods are different, or there was a developmental trend in the data (the groups and the time periods were not independent of each other). Finally, I shall indicate the overall emphasis placed by the two groups taken as a whole on any given variable. If the approximate average arithmetical mean score for the three groups taken together is within the upper third of the range of possible scores on the variable, it will be called high emphasis; if the average falls in the middle third of the range, it will be called moderate emphasis; if the score falls in the bottom third, low emphasis. I shall test the research hypothesis about the per formance of each variable in relation to each of the two groups. I shall test the second hypothesis, and so forth, until I have tested all the hypotheses under the heading. I shall then offer an overall summary of the findings. In some instances, this will be very brief and in others, more detailed. !! ; I 195 Expectations of Leaders The five variables reported here are measures of different norms. Each measures a specific type of generalized prescription applied to all members of an organization, regard less of position. Taken together, these norms are a measure 2 of organizational climate. More specifically, they are a measure of a particular aspect of organizational climate: the norms associated with the structure of authority in the organization. Given the nature of the environment surrounding the three field offices involved in the study and the change strategy employed in the experimental office, the following general research hypothesis was formulated: The experiment in organizational devolution will enable the authority structure of the experimental organization to change toward a multi-hierarchical, collaborative structure. This change is expected to be indicated by the lead ership expectation variables in the following ways: 1. A significant increase in rules orientation. The experimental conditions gave the organization the opportunity to create many of its own rules and policies. Consequently, members of the organization are expected to place greater emphasis on rules and policies. 2. A significant increase innurturance of subordinates. Since all normally required reports and evaluative procedures : were suspended during the experiment supervisors are expected to spend more time with their subordinates. In general, this increased interaction is expected to increase the importance I I 196 that subordinates attribute to feeling that their superiors care about their well-being. 3. A significant decrease in closeness of supervision. To the extent supervisors and subordinates paid more attention to each other, each would probably become more aware of the needs and skills of the other and adapt his or her behavior accordingly. This is expected to decrease the demand for close supervision. 4. A significant decrease in universalism. To the extent the second and third results occurred, members of the organization would grow to expect equal and fair treatment and consideration of unique needs; thus, a decline in universalism is expected. 5. An increase in expectations of promotion and achievement. Because of the experiment, it would be possible for members of the organization to explore many novel ways of doing things and to develop new skills; consequently, the demand for promotional and achieve ment opportunities is expected to increase. Rules Orientation A special notation will be used to present the data. The experimental, aware control and unaware control groups associated with the block of data called All Respondents will be denoted by El, Al, and UA1, respectively. The experimental, aware control, and unaware control groups associated with the data called Four-Time Respondents will be denoted by E2, A2, and UA2. The predicted increase in rules orientation did not occur in either El or E2 (Tables III & IV). There were no significant differences between El, Al and UA1 through the 197 c i I d Mean Scores by Organization per Time Period " FalT.. Spring Fall Summer 1967 1968 1968 1969 20 ■ ^---- > 1 Two-Way Analysis of Variance: All Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 23.407 255 Between 25.582 11 A 47.805 2 2.0509 0.1286 B 28.847 3 1.2376 0.2960 AB 16.542 6 0.7097 0.6443 Within 23.309 244 aN = 22 per organization per time period = experimental; A = aware control; UA = unaware control TABLE III VARIABLE 1: RULES ORIENTATION 198 ci k) Mean Scores by Organization cer Time Period Fall 1968 Fall 1957 Spring 1968 20 OE Groups by Trials Analysis of Variance: Four-Time Respondents Source Mean Squares Degrees of Freedom F- Patio ProbabiJ ity Total 24.7122 95 Between 74.5177 23 Groups 38.0000 2 0.487 0.6263 Error(G) 77.9955 21 Within 8.8021 72 Trials 4.5659 3 0.474 0.7056 G by T 2.1806 6 0.226 0.9656 Error(T) 9.6344 63 aN ~ 8 per organization per time period = experimental; A = aware control; UA = unaware control TABLE IV VARIABLE 1: RULES ORIENTATION 199 two-year period; moreover, there were no significant differences in the aggregate data gathered during each of the time periods; and there were no developmental trends in the data. The same set of relationships held for groups E2, A2, and UA2. Both groups placed only a moderate emphasis on the 3 norm of rules orientation. The experiment had no discernible effect on this variable, flatly refuting the hypothesis that emphasis on rules would increase during the study. Nurturance of Subordinates The expectation of nurturance of subordinates increased significantly (p=.007), and in the direction predicted by the research hypothesis, in group El. This change occurred between the Fall of 1967 and the Spring of 1968; after the Spring of 1968, nurturance remained relatively constant. (Table V) Taken as a whole, the data gathered from the All Respondents group did not vary significantly during the two years. There were no significant differences in the data from each of the time periods and there were no develop mental trends. Both the All Respondents group and the Four-Time Respondents group placed a relatively high emphasis on 4 nurturance of subordinates. (Table VI) 200 Mean Scores9 by Organization*3 per Time Period Fall- 1967 Fall 1968 30 20 Two-■Way Analysis of Variance: All Respondents Source Mean Squares Degrees of _ _ , . Freedom . ^-Ratio Probability Total 16.152 255 Between 30.630 11 A 79.011 2 5.0978 0.0070 B 19.532 3 1.2602 0.2879 AB 20.052 6 1.2938 0.2597 Within 15.499 244 aN = 22 per organization per time period ^E = experimental; A = aware control; UA = unaware control TABLE V VARIABLE 2: NURTURANCE OF SUBORDINATES 201 Mean Scoresa by Organization*5 per Time Period Fall 1968 pring 1968 30 UA 20 Group:; by Trials Analysis of Variance: Four--Time Respondents Source Mean Squares Degrees of F reedom f -Ratio Probability Total 14.4838 95 Between 21.8460 23 Groups 16.7607 2 0.751 0.5117 Error(G) 22.3304 21 Within 12.1319 72 Trials 7.0697 3 0.598 0.6224 G by T 17.9964 6 1.523 0.1845 Error(T) 11.8145 63 aN = 8 per organization per time period ^E = experimental; A = aware control; UA’-- unaware control TABLE VI VARIABLE 2: NURTURANCE OF SUBORDINATES Closeness of Supervision Contrary to the research hypothesis, the expectation of close supervision increased in both experimental groups (El; p=.0001: E2; p=.001) and decreased in the respective control groups (Tables VII & VIII). In the All Respondents population, there were no differences in the data from the several time periods in the control groups expectation decreased and there was a significant downward trend in the overall data (p=.05). Data collected from the Four-Time Respondents population were not significantly different when the time periods were compared and no overall trend in the data was revealed over the two-year period. Both the All Respondents population and the Four-Time Respondent population placed moderately low emphasis on closeness of supervision.^ Universalism The research hypothesis predicted that the conventional meaning of universalism, equal standards applied to all, would be transformed by the experiment. The scale used to measure the norm of universalism presupposed a continuum which ranged from totally non-universalistic (totally situation specific) behavior to totally universalistic stand ards applied rigidly to all members of the organization. The research hypothesis predicted the emergence of a more flexible form of universalism; people would expect to 203 Mean Scores by Organization per Time Period Fall 1967 Fall 1968 Spring 1968 20 10 Two-Way Analysis of Variance: All Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 15.276 255 Between 55.271 11 A 207.036 2 15.3674 0.0000 B 7.491 3 0.5560 0.6489 AB 28.573 6 2.1208 0.0510 Within 13.472 244 aN ~ 22 per organization per time period = experimental; A = aware control; UA = unaware control TABLE VII VARIABLE 3: CLOSENESS OF SUPERVISION I 204 Mean Scores3 by Organization^ per Time Period Spring 1963 Fall 1968 Fall 1957 Summer 1969 20 10 Groups by Trials Analysis of Variance: Four-Time Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 17.3895 95 Between 44.5000 23 Groups 244.3438 2 9.594 0.0014 Error(G) 25.4673 21 V7it.hin 8.7292 72 Trials 10.7500 3 1.253 0.2976 G by T 9.3021 6 1.084 0.3816 Error (T) 8.5784 63 aN = 8 per organization per time period E = experimental; A = aware control; UA = unaware- control TABLE VIII VARIABLE 3: CLOSENESS OF SUPERVISION 205 be treated equally and according to specific individual needs. The measurement technique was not designed to allow for the identification of this kind of universalism; consequently, the research hypothesis predicted that the new meaning of universalism would be reflected by a decreasing score on the scale. This hypothesis was not supported by the data. Group El demanded greater universalism as a consequence of organi zational devolution (p=.01). (Table IX) Change occurred between the Fall of 1967 and the Spring of 1968 and no dis cernible changes occurred in the periods that followed. Taken as a whole, the data gathered from the All Respondents popu lation did not vary significantly during the several time periods. No developmental trends could be identified. Data taken from the Four-Time Respondents population did not reveal any changes of any kind. There was no significant change in the demand for universalism among the groups E2, A2 and UA2 during the study, the four blocks of data were not dissimilar and no trends were discerned. (Cf. Table X) Both sets of groups placed moderate emphasis on equal treatment of all personnel in the organization.^ The experiment in organizational devolution seems to have had some effect on the norm of universalism; this was indicated by a significant increase in expectations associated with this norm in the All Respondents group; no such change 206 Mean Scores by Organization per Tine Period Fall 1967 Fall 1968 prmg 1968 20 10 Two-Way /analysis of Variance: All Burpondonts Source Mean Squares Degrees of Freedom F- Ratio Probability Total 12.652 255 Between 25.848 11 A 53.968 2 4.4762 0.0123 B 10.352 3 0.8586 0.5342 AB 24.222 6 2.0090 0.0645 Within 12.057 244 aN = 22 per organization per tine period = experimental; A = aware control; UA = unaware ’control TABLE IX VARIABLE 4: UNIVERSALISM 207 3 b Mean Scores by Organization per Time Period rail 19S7 Fall 1968 prang 1968 20- 1C- Groups by Trials Analysis of Variance: Four-Time Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 12.5788 95 Between 20.2061 23 Groups 35.0730 2 1.867 0.1780 Error(G) 18.7902 21 Within 10.1424 72 Trials 3.5938 3 0.324 0.8103 G by T 3.4063 6 0.307 0.9305 Error(T) 11.0957 63 aN = 8 per organization per time period = experimental; A = aware control; UA = unaware control TABLE X VARIABLE 4: UNIVERSALISM 208 occurred in the Four-Time Respondents group. The research hypothesis was completely inappropriate; it assumed the norm was strong before the experiment and would be transformed during the actual course of the experiment. In fact, the norm was only moderately strong and was affected only in a relatively minor way. Promotion-Achievement Because the experiment was likely to afford the members of the experimental organization increased opportunities to develop new skills and new behavior patterns, the research hypothesis predicted an increase in the promotion-achievement norm. This prediction was fulfilled by the El group (p-.0009). As before, the change in expectations that produced this finding occurred between the first and second times that data were gathered. The aggregate blocks of data collected from the All Respondents population did not vary significantly and there was no significant tendency for the three groups to change as a function of time. (Table XI). Findings for the second general group, the Four-Time Respondents, did not support the research hypothesis. No significant results were produced by the three statistical tests. (Table XII) Both the All Respondents population and the Four-Time Respondents placed moderately high emphasis on promotion- achievement . ^ 209 cl b Mean Scores by Organization per Time Period i’ alT"" Spring Iall Sumner 1967 1968 1968 1969 20 □---- --= 0 = — 10 A---- = § ea Two-Way Analysis of Variance: All Respondents Source Mean Squnres degrees of Freedom F-Patio Probability Total Between A B AB Within 13.871 27.361 101.413 6.298 13.209 13.262 255 11 2 3 6 244 7.6467 0.4748 0.9960 0.0009 0.7043 0.5707 aN = 22 per organization per time period E = experimental; A = aware control; UA = unaware control TABLE XI VARIABLE 5: PROMOTION-ACHIEVEMENT 210 Mean Scores3 by Organization^ per Time Period Fall 19S7 Fall 1968 Spring 1968' Sumner 1969 20 10 Groups by Trials Analysis of Variance: Four-Time Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 10.2522 95 - Betv.’ een 14.7591 23 Groups 8.2605 2 0.537 0.5972 Error(G) 15.3780 21 Within 8.8125 72 Trials 2.2917 3 0.262 0.8533 G by T 12.8854 6 1.475 0.2005 Error(T) 8.7351 63 aN = 8 per organization per time period = experimental; A = aware control; UA = unaware control TABLE XII VARIABLE 5: PROMOTION-ACHIEVEMENT 211 In this case the research hypothesis was partially supported by the data; the All Respondents group (El) increased its expectation of promotion and achievement. The Four-Time Respondents group in the experimental organization did not change as a consequence of organizational devolution. Summary This section has explored the effects of organizational devolution on the authority structure of the experimental organization. Measurement of this structure was obtained through the use of five generalized organizational norms: rules orientation, nuturance of subordinates, closeness of supervision, universalism and promotion-achievement. The experiment seems to have had remarkably different effects on the two groups that were analyzed. The group called All Respondents is described in considerable detail in Chapter V; the chief characteristic of this group is that its membership changed. Old members dropped out and new members were added. Consequently, the statistical analysis applied to this data was one which would tend to be conservative in estimating change; it was based on the assumption that the data gathered from each period were independent of the data gathered from the other periods. Because of this assumption it takes a sharper degree of change to produce a statistically significant outcome. Despite this, the All Respondents group was strongly affected by the 212 experiment; significant changes occurred on four out of five measures taken from this group. These changes were as follows: 1) the nurturanceof subordinates norm became stronger, 2) the demand for close supervision increased, 3) the demand for universalistic standards increased, and 4) the demand for promotion and achievement opportunities increased. No change occurred in rules orientation. The results of the experiment had very different effects on the Four-Time Respondents. The general level of measurement obtained from this group and the All Respondents group was quite comparable on all five variables. Mean scores for the Four-Time Respondents group were consistently slightly lower overall on each variable than mean scores for the All Respondents group. The statistical analysis applied to this data was well suited. It did not require that the data from each time period were independent of the data from other time periods, and it is more sensitive to changes than was the statistical analysis applied to the All Respondents group. Despite these conditions, the Four-Time Respondents were remarkably unaffected by the experiment. On four out of five measures, no significant changes occurred. The norms of rules orientation,nurturance of subordinates, universalism and promotion-achievement were not affected by the experiment. This group did change its expectations about close supervision. This change could be graphed as an inverted "U". Starting with the first time period, the norm increased. It stayed 213 about the same during the second and third periods and decreased during the last. The research hypotheses associated with the five variables were supported twice by the All Respondents group and received no support from the Four-Time Respondents group. For the All Respondents group, the research hypothesis pre dicted that rules orientation would increase and it was unchanged. A second hypothesis predicted that closeness of supervision would decrease and it increased. A third hypo thesis predicted that universalism would decrease and it increased. In all instances, the research hypotheses were not supported by the Four-Time Respondents group; in four instances the hypotheses predicted change and no change occurred. In one instance, change was predicted and occurred but it was in the opposite direction from that predicted. Task Characteristics In other words, these variables are a product of the way organizational tasks are defined by the general norms of the organization and the way the individual feels about himself at work. In Chapter I, the role that these variables play in defining the nature of an authority structure was described. Put briefly, job tension is an anxiety about being unable to do the job, or not being sure how to do it, or both. Participation in decisions is a measure of how authority is exercised; the greater the general level of participation, 214 the more collaborative the authority structure; the lower the level of participation the more the authority structure is based on dominative power. The variables included in this section and the infor mation they provide about the effects of the experiment are part of the test of the general research hypothesis. This hypothesis predicts a movement toward a new pattern of authority in the experimental office. In terms of the three variables related to the character of tasks in the organization, the following changes are predicted: 1. An increased level of participation in decisions. 2. A decrease in job tension. 3. More positive attitudes toward work-related changes. Participation in Decisions Although the same general kind of change in the data occurred with this variable that occurred with several other variables, this change was not great enough to be statistically significant. The All Respondents population did not indicate that participation had increased as a result of the experiment. (Table XIII) The aggregate data gathered during each period remained comparable and there were no trends simultaneously affecting all three groups (El, Al, and UA1) through time. The Four-Time Respondents also indicated that participation in decisions was not effected by the strategy of organizational devolution, agreeing with the All Respondents 215 Mean Scores by Organization per Time Period TaTT 1968 Spring 1968 Summer 1969 30 20 - Two-Way Analys is of Variance: All Respondents Source Mean Squares Degrees of Freedom F-- Ratio Probability Total 27.218 255 Between 8.755 11 A 14.799 2 0.5276 0.5964 B 1.573 3 0.0561 0.9818 AB 10.332 6 0.3683 0.8986 Within 28.050 244 a N = 22 per organization per time period • E = experimental; A = aware control; UA = unaware control TABLE XIII VARIABLE 6: PARTICIPATION IN DECISIONS i l ; ! 216 Mean Scores by Organization per Time Period Fall Spring Fall Summer 1967 1968 1968 1969 30- - 20 —TV -o— " ^ q S'ua Groups by Trials Analysis of Variance: Four-Time Respondents Source Mean Squares Degrees of F reedom F-Ratio Probability Total 19.2206 95 Between 37.6069 23 Groups 10.0417 2 0.250 0.7841 Error(G) 40.2321 21 Within 13.3472 72 Trials 98.1250 3 10.597 0.0001 G by T 13.8750 6 1.498 0.1926 Error(T) 9.2599 63 aN = 8 per organization per time period = experimental; A = aware control; UA - - unaware control TABLE XIV VARIABLE 6: PARTICIPATION IN DECISIONS 217 group. There was a significant difference in the aggregate data gathered during each of the time periods; periods Tl and T2 were relatively comparable and periods T3 and T4 were comparable but the scores from the latter two periods were significantly higher than from the former periods (p=.0001). (Table XIV). There were no overall trends simultaneously affecting all three organizations, despite the above finding. Closer observation of the data shows that there was no consistent pattern of change common to E2, A2 and UA2. The amount of emphasis placed on this variable was 8 moderately high. Job Tension The research hypothesis was not supported by the All Respondents group; no change occurred in the amount of job tension experienced. And, the aggregate data was comparable across the time periods. No trends were revealed in the data. (Table XV) The same conditions held for the Four-Time Respondents. (Table XVI) 9 The amount of emphasis placed on job tension was low. Work-Related Change Attitudes toward work-related change remained stable during the period of the study in both the All Respondents and the Four-Time Respondents groups. (Tables XVII and XVIII) The four blocks of aggregate data remained comparable in 218 Mean Scores by Organization per Time Period i all 1968 Pall 1967 30 Two-Way Analy.sis of Variance; All P.e-*.onuents Source .'lean Squares Degrees of Freedom F-Katio rrebabi lxt:y Total 49.797 255 Between 23.823 11 A 43.305 2 0.8496 0.5679 B 6.698 3 0.1314 0.9406 AB 25.891 6 0.5080 0.8035 Within 50.968 244 N = 22 per organization per time period E = experimental; A = aware control; UA = unaware control TABLE XV VARIABLE 7; JOB TENSION 219 Mean Score.sa by Organization*5 per Time Period Fall 19S7 Fall 1968 Spring 1968' *0= 30 — Groups by Trials Analysis of Variance: Four-Time Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 42.5131 95 Between 107.4778 23 Groups 143.1357 2 1.375 0.2741 Error(G) 104.0818 21 Within 21.7604 72 Trials 37.9551 3 1.857 0.1446 G by T 27.5798 6 1.350 0.2482 Error(T) 20.4350 63 aN = 8 per organization per time period - experimental; A = aware control; UA = unaware control TABLE XVI VARIABLE 7: JOB TENSION 220 Mean Scores by Organization per Time Period FaTT 1967 Fall 1968 Spring 1969 30 7v;o-Way Analy ,is of Variance: All P.esj.ondents Source Mean Squares Degrees of Freedom P-Ratio Probability Total Between A B AB Within 29.097 23.596 13.492 1.232 38.146 29.346 255 11 2 3 6 244 0.4597 0.0420 1.2999 0.6379 0.9879 0.2569 a N = 22 per organization per time pc>riod experimental; A = aware control; UA = unaware control TABLE XVII VARIABLE 8: WORK—RELATED CHANGE 221 c l I d Mean Scores by Organization per Time Period Fall Spring Fall Summer 1957 1968 1968 1969 30 - & » < ^ 6 a - "0A Groups by Trials Analysis of Variance: Four-Tine Respondents Source Mean Squares Degrees of Freedom F-Ratio Probability Total 23.2631 95 Between 67.2061 23 Groups 11.6982 2 0.161 0.8525 Error(G) 72.4926 21 Within 9.2257 72 Trials 8.6771 3 0.938 0.5702 G by T 9.1979 6 0.994 0.5619 Error(T) 9.2545 63 *N = 8 per organization per time period = experimental; A = aware control; UA= unaware control TABLE XVIII VARIABLE 8: WORK-RELATED CHANGE 222 both groups and there were no trends in the data over time. Emphasis on work-related change was moderate.^ Summary The research hypotheses were refuted by both groups on all three variables— participation in decisions, job tension and work-related change. The experiment had no effects upon task characteristics. Self at Work In this last section of the research findings, the organizational devolution experiment will be explored in terms of its effect it had on how individuals felt about themselves while at work. Two variables are involved in this analysis: acceptance of self and acceptance of others. These variables indicate how much a person accepts his reactions, feelings, attitudes and beliefs as his or her own and as natural and is willing to accept similar features of others in the same manner. The research hypotheses predicted that: 1. Acceptance of the self would increase in the experimental organization. 2. Acceptance of others would increase. Self-Acceptance The predicted increase in self-acceptance did not occur in either the All Respondent or the Four-Time Res pondents groups. Data from the four time periods were 223 comparable and there were no trends affecting all three groups at the same time; these conditions held for both the All Respondents and the Four-Time Respondents groups. Emphasis on self-acceptance was high.^ (Tables XIX & XX) Acceptance of Others Again, the research hypothesis was refuted by the data provided by both general populations (Tables XXI & XXII) The experimental, aware control and unaware control were not different from each other for both the All Respondents and the Four-Time Respondents groups. Data collected from these two groups during the study were comparable regardless of time period. There were no trends affecting the aggregate populations over time. 12 Emphasis on acceptance of others was moderate. Summary Both variables reported in this section remained unchanged in every way during the two years that data were gathered. Devolution of authority to the field office had no effect upon how individuals who worked for the organization accepted themselves, or accepted others. 224 Mean Scores" by Organization per Time Period Fail 150 l96V Fail 1968 1968 140 130 120 , T1. . - J . .* ^ -Wa* Analvs: _s of '/arian: -*! All Fes'-r nder.ts Scarce Man Sousres Degrees of Freedom ?-Katie Probability Total 484.601 255 Between 590.993 11 A 749.668 2 1.5624 0.2099 B 488.764 3 1.0187 0.3860 AB 589.215 6 1.2280 0.2916 Within 479.805 244 N = 22 per organization per time period ^E - experimental; A = aware control; UA = unaware control TABLE XIX VARIABLE 9: SELF-ACCEPTANCE 225- Mean Scores by Organization per Time Period Fall 1967 Spring 1968 Fall 1968 150 140 130 120 Source Mean Squares Degrees of Freedom F-Ratio Probability Total 631.8289 95 Between 1446.0435 23 Groups 1786.2969 2 1.264 0.3032 Error(G) 1413.6384 21 Within 371.7326 72 Trials 77.1250 3 0.205 0.8923 G by T 480.3177 6 1.279 0.2790 Error(T) 375.4201 63 N = 8 per organization per time period 5E = experimental; A = aware control; UA = unaware control TABLE XX VARIABLE 9: SELF-ACCEPTANCE 226 Mean Scores* by Organization per Tine Period Fall 1967 Fall 1963 120 110 100 70 Two-Way /.nalysrs of Variance : All Re spend on ts Source Mean Sauares Degrees of Freedom F-Ratio Probability Total 135.079 255 Between 91.457 11 A 273.299 2 1.9942 0.1361 B 21.077 3 0.1538 0.9266 A13 66.034 6 0.4818 0.8228 Within 137.046 244 a.. N =■ 22 per organization per time period experimental; A = aware control; UA = unaware control TABLE XXI VARIABLE 10: ACCEPTANCE OF OTHERS 227 Mean Scores by Organization per Tine Period Fall 1567 196 120 110 100 90 Two -Wa y An a 1 y s 5 s of Variance: Four-Tine fRespondents Source Fear. Squares Degrees of Freedom F-Ratio Probability Total 192.1891 95 Between 433.8244 23 Groups 591.2617 2 1.412 0.2653 Error(G) 418.8304 21 Within 115.0000 72 Trials 177.0703 3 1.480 0.2273 G by T 35.3711 6 0.296 0.9361 Error(T) 119.6280 63 aN *= 8 per organization per time period - experimental; A = aware control; UA = unaware control TABLE XXII VARIABLE 10: ACCEPTANCE OF OTHERS 228 Summary Statistical analysis and interpreation of the research data revealed a striking lack of change in the experimental organization. Two groups were involved in this analysis: All Respondents and Four-Time Respondents. On the ten measures applied to the All Respondents group, four instances of change were found; all were associated with leadership expectations. Nurturnace of subordinates grew stronger, the demand for close supervision increased, the demand for universalistic treatment of employees increased, and expectations of promotion and achievement increased. No change occurred in the norm of rules orientation. The Four-Time Respondents group changed in a way that could be attributed to the experiment in only one area— closeness of supervision. It is noteworthy that these changes all occurred during the same general interval of time: between the Fall of 1967 and the Spring of 1968. During this period several important events occurred. A new leader was appointed in the district office and a team building effort was undertaken. Following this period, all the changes stopped for the 13 rest of the experiment. In the other areas of measurement, task characteristics which included three variables— participation in decisions. 229 job tension, and work-related change— and self-at-work which included measures of self-acceptance and acceptance of others, no change occurred in either the All Respondents or the Four- Time Respondents groups. There was neither a significant 14 increase or decrease in these variables. Overall, these data could be dismaying if the research hypotheses were meaningful. In retrospect, the research hypotheses were far too bold. Let us pause and reflect on the nature of the experi ment in organizational devolution. All formal relationships with the overall SCIF organization were severed; the field office reported only to the General Manager and this did not mean meeting the General Manager's expectations but simply making him aware of what was happening in the field office. In fact, these reports were few and very far between. All the normal reports, evaluation requirements and support services flowing from the SCIF to the field offices were suspended. In short, the field office was, to the maximum extent possible, put on its own. Given the normal assumptions in organization theory about the role of hierarchy and the necessity for it, a conventional prediction would have been of absolute chaos, of no work being accomplished, of the morale of the organization being destroyed, and of other catastrophes. 230 What the research data reveal is that without the external hierarchy, the field office operated more or less the same as it always had except that there was more concern for the well-being of members of the organization, an increased demand for supervision, an increase in the demand that all members of the organization be treated equally and greater opportunities for promotion and achievement were expected. For all practical purposes, the hierarchy normally imposed on the field office was unnecessary. The field office was perfectly able to decide what kinds of help it needed, obtain the services, and continue to function, on its own without centralized direction and coordination. NOTES CHAPTER VI 1. Donald J. Veldman, Fortran Programming for the Behavioral Sciences (New York: Holt, Rinehart and Winston, r?S7)','"p'.'257.------ 2. Robert Kahn, et al., Organizational Stress (New York: John Wiley and Sons, Inc., 1964), p.150. 3. Average score for each of the six groups was approximately 19. Maximum possible score that could be obtained was 30. 4. Average score approximately 26; maximum possible score was 30. 5. Average score approximately 13; maximum possible score was 30. 6. Average score was approximately 13; maximum possible score was 30. 7. Average score approximately 19; maximum possible score was 30. 8. Average score approximately 22; maximum possible score was 30. 9. Average score approximately 30; maximum possible was 90. 10. Average score approximately 30; maximum possible was 45. 11. Average score was approximately 135; maximum possible was 160. 12. Average score approximately 98; maximum possible score was 135. 13. Cf. Chapter IV. 14. Mean scores for all groups and all variables can be found in Appendix VI. 231 CHAPTER VII SUMMARY AND CONCLUSIONS I began this report with a discussion of the need for experimentation in the design of formal organizations and identified a series of issues in the field of organization development and organization design. I gave considerable attention to the issue of the relationship between organi zational environments and organizational performance. Current theory and research suggest that environments are an important determinant of organizational effectiveness and can induce or promote change, or both, in the structures of authority of organizations. While this issue has been raised in the literature, there has been little improvement in the underlying theory. In some instances, this has led to an oversimplification of the complexity of change and an underestimation of the variety of new organizational designs that may emerge. In particular, there has been speculation that environmental turbulence will help to create — even force — the emergence of non- bureaucratic, non-hierarchical organizations. I argued that non-bureaucratic organizations were unlikely to take only one form (non-hierarchical). Multi-hierarchical, collaborative organizations were discussed as one alternative. 232 233 Contemporary strategies of organization development (OD) that developed roughly during the same period that the question of organizational environments began to appear were discussed. I noted that the two OD strategies share a common concern with the structure of authority of organizations and both are concerned with the problem of organizational effectiveness. Both OD strategies, Grid Organization Development and Emergent/Situational Intervention Theory, approach organi zational change by focusing on the behavioral processes of an organization; both assume that if the processes of an organi zation change, the overall performance will change. I described two models of organizational processes. This material was based on the work of Herbert Shepard;*1 who shows that what he calls the primary principles of organization (traditional principles of organization) and the secondary (non-bureaucratic) principles are related to each other. The secondary system emerges from the primary and does not exist completely to the exclusion of it. Shepard says that the primary processes of organization can give way to the secondary under appropriate circumstances and vice versa. Although Shepard's theory is interesting, it heretofore has never been subjected to empirical testing. I noted that this report would test some of these ideas. The report would assess the effects of a new organi zation development intervention— organizational devolution, 234 in a public agency of the State of California, the State Compensation Insurance Fund (SCIF). A brief discussion of organizational devolution followed and I presented the research question that guided the study. This question was: Did the experiment in organizational devolution promote the development of a multi-hierarchical, collaborative structure of authority in the experimental organization? Since Shepard's theory did not provide a precise defin ition of authority structure, it was necessary to pursue this problem at some length. I offered a definition of authority structures, formulated in such a way as to be applicable to a variety of different organizations. Ten operational definitions and indices were described. These ten indices or research variables will be used to assess the effects of organizational devolution in a field office of the SCIF. I presented ten research hypotheses related to these variables. In an effort to identify more precisely the character istics of the organizational development program in SCIF and to contribute to a more precise definition of other OD efforts, Chapter II re-introduced the issue of organizational environ ments and organizational design. A new formulation, based on 2 previous work by Emery and Tnst. was proposed. This formu lation used four variables to define environments; four environments— Early Industrial, Industrial, Late Industrial, and Post-Industrial— were discussed at length. Then, a similar procedure was used to describe four organizational designs— Pre-Bureaucratic, Centralized Bureaucratic, 235 Decentralized Bureaucratic, and Non-Bureaucratic— which were theoretically congruent with the four environments. I gave special attention to the non-bureaucratic design, in cluding a discussion of criteria of organizational effective ness and an example of one such design— the Consociated Model. A brief history of SCIF followed; this history paid considerable attention to the design of the organization during several periods in the past and to the design used until a few years ago. When the last organization design was described, forces from the external environment, and from inside the organization that were demanding change, were detailed. In response to these forces an OD program was begun; one consequence of this program was the experiment in organi zational devolution undertaken in the field office which is the subject of this report. Prior to Chapter III, OD had not been rigorously de fined. More precise definition of OD was the concern of early parts of this chapter. In the process of clarifying the definition of OD, the first step was to distinguish management development from organization development. Then I presented and criticized five definitions of OD available in the literature and proposed a new formulation. This formulation distinguished organization development from organizational change. A common feature to the definitions of OD in litera ture and the one proposed is that OD is oriented to an organi zation as a total system and is concerned about overall system 236 effectiveness. The two major strategies of OD highlighted in Chapter I were now described in much fuller detail. I explained Grid OD, the nature of its interventions— instrumented laboratories— and the phasing of its interventions. I reviewed three major research evaluations of the applications of this approach [a plastics firm (Sigma), the U.S. Forest Service, and the Internal Revenue Service]. I examined Emergent/Situational Intervention Theory (E/S) in a similar way and several of the interventions associated with this theory— team building, intergroup building, deep sensing, and organizational mirror— were described. Three major OD research evaluations were described and criticized. Two represented examples of the application of E/S Theory (the Bank and the Engineering firm); one did not represent either E/S or Grid OD but could not be ignored in the review of major OD evaluations. I discussed, under the heading of E/S Theory, the intervention called Organizational Devolution. I presented and reformulated the theory behind this intervention. During this analysis I noted that the concept of devolution which describes non-hierarchical organizational structures was difficult to explain because of the pervasive role the concept of hierarchy has played in organization theory and the re sulting lack of concepts to describe non-hierarchical relation ships. I detailed the exact form that the organization devolution intervention took in the SCIF experiment. Chapter IV offered a report of the experiment written from the viewpoint of a participant observer. Events that occurred just prior to starting the experiment formally, the strategy followed by the District Manager during the early moments of the experiment, the appointment of a new District Manager, and the additional events that seemed to have shaped the two-year experiment were highlighted. Chapter V set forth the research design, data- gathering strategy and statistical analysis. The research design is a Longitudinal, berfore and After Design with Two Control Groups. The strengths and weaknesses of the design were explained in terms of the extent to which it controlled nine threats to internal validity, i.e., threats to unambi- gous interpretation of the data. The data-gathering strategy, including the method used to construct the 10 attitudinal scales, were analyzed in terms of potential undesirable- consequences. This analysis was based on a general criticism of the behavioral sciences developed by Chris Argyris in Intervention Theory and Method. I warned the reader to bear in mind these criticisms when the research findings were presented. Chapter VI presented the research findings. This material was organized into three areas of investigation: Leadership Expectations, Task Characteristics, and Self at Work. The first of these areas measured generalized norms 238 associated with the authority structure of the organization and was a partial assessment of organizational climate. The second area, task characteristics, measured aspects of the tasks which were defined by the authority structure. The last area, Self at Work, explored how people felt about them selves in the context of the authority structure. The research findings in each of these areas are summarized in Fig. 9 below. Of the ten measures applied to the All Respondents Group, there was significant change in four variables: Nurturance of Subordinates, Closeness of Supervision, Universal- ism, and Promotion-Achievement. In all four cases, the changes were in the direction of stronger norms. The Four-Time Respondents produced only one significant change in attitudes: increasing expectations of close supervision. In all other areas and on all other variables, no significant change occurred. The research hypotheses were supported only twice. The variables nurturance of Subordinates and Promotion- Achievement, increased as predicted. In all other cases the hypotheses were refuted. At the same time, these findings show that during the experiment the organization, with the exceptions of demanding greater consideration of employees (nurturance of Subordinates) and increasing expectations of Promotion and Achievement, continued to function as it had when it was part of the hierarchy of the SCIF before the experiment. In short, the j l 239 Experimental Groups AREA All Respondents Four-Time Respondents A. Expectations of Leaders 1. Rules Orien tation No Change No Change 2. Nurturance of Subordinates Significant Increase No Change 3. Closeness of Supervision Significant Increase Significant Increase 4. Universalism Significant Increase No Change 5. Promotion- Achievement Significant Increase No Change B. Task Character istics 6. Participation in Decisions No Change No Change 7. Job Tension No Change No Change 8. Work-Related Change No Change No Change C. Self at Work 9. Self-Acceptance No Change No Change 10. Acceptance of Others No Change No Change Fig. 9. Summary of research findings j ! 240 field office demonstrated that it was possible to maintain an authority structure in the field office without having the traditional external organizational hierarchy. Conclusion and Future Directions At the end of the two-year experiment in organizational devolution, top management of SCIF recognized what this study has confirmed, that the field office had been able to operate on its own without hierarchical direction and control. Con sequently, top management decided to provide the same condi tions of autonomy to all field offices and undertake a major effort to create a non-bureacratic public organization. 4 The organizational design which was envisioned and which is now well toward being established is portrayed in Fig. 10. I shall point out several of the more interesting features of this organization chart. The fact that it identi fies two types of environments— focused and diffused— is a tribute to the importance placed on relating to the environ ment. The focused environment represents those aspects of the society that relate directly to the SCIF operations; the diffusedenvironment represents the larger cultural setting. The blocks on either side and at the top of the chart repre sent the three policy formulating boards of the organization. These groups are interdependent with each other; no one group is in a position to dominate the other. Collaboration is 241 Diffused Environment Focused Environment District Mgr's Council (23) o H hh ( P Hi 10 H - H- O d < D O *0 3 IQ Staff Services \/ ft Hi Training H- (10) Personnel (23) Finance Data Process Programs Field Offices (23) ^ > '/I Fig. 10.— SCIF organization chart 242 required for decisions. The Office of the President incor porates all upper middle and top managers from the old organi zational structure; these persons are expected to contribute to the formation of policy experience and sensitivity of the economic and political environment. The District Manager's Council is made up of all managers of Field Offices; they are expected to contribute to the formation of policy their awareness of the needs and problems of this most crucial sub system of the organization. The Program Advisory Council is composed of two elected, non-supervisory persons from each field office. They are expected to contribute a sound know- 5 ledge of the problems the organization faces with its clients. The four rectangles in the center of the chart, Staff Services represent the traditional categories of staff ser vices in organizations like SCIF, but placed in an entirely novel role. These units depend on the rest of the organization for their survival; either the field offices or the policy making boards can contract with the service units for specific research and help. If the service is not adequate, the field offices and policy boards, to the extent possible under California law, can get services elsewhere. If this occurs to a great extent, the service units could "go out of business.v All sub-systems of the organization have been placed on a decentralized budget format similar to what was used with the field office in the experiment (Chapter III). Pro grams common to several field offices are collected and know- 243 ledge of success or failure in reaching objectives is shared with all offices. Feedback on these and other programs is in terms of the goals that the field offices set for themselves and were accepted by representatives of the three policy making boards at the beginning of the fiscal year. The goal of the organization design that is currently being implemented is to maximize the opportunity for adaptive ness and effectiveness and to continue to take advantage of what was learned from the original experiment: centralized, hierarchical command is not necessary. NOTES CHAPTER VII 1. Herbert Shepard, "Changing Interpersonal and Group Processes," in Handbook of Organizations. Edited by James March (Chicago: Rand McNally & Company, 1965), 115-43. 2. E. F. Emery and E. L. Trist, "The Causal Texture of Organizational Environments," Human Relations, XVII (February, 1965), 21-37. 3. Chris Argyris, Intervention Theory and Method (Reading, Mass.: Addison-Wesley Publishing Company, 1970). 4. Neely Gardner, "Organizing for the Future— a Trial." (Paper presented at the 1971 National Conference of the American Society for Public Administration, Los Angeles, Calif., University of Southern California School of Public Administration, May, 1971, Mimeographed, 11 pp.). 5. Cf. Appendix 5. APPENDIX I ! ! PROGRAM STATEMENT PART I South Pasadena District Office State Compensation Insurance Fund To achieve and maintain leadership in the workmen's compen sation insurance industry, the State Fund must find the organ ization and management style best for its effective operation and successful achievement of its role within the social insurance system. Leadership requires equity to the insured employer, his employees and Fund personnel. Constructive participation in the rate making and classification process, benefit level determination, a sound fiscal condition and a level of service assuring effective competitive operations are also demanded. This experiment in organizational structure and management style is designed to test one method of achieving and maintain ing leadership within the industry and meeting the Fund's Social objectives. The program of the South Pasadena Office is geared to the total programs of the State Compensation Insurance Fund. These programs are carried on to meet a social need for Work men' s compensation insurance — insurance that was not compul sory prior to the founding of the State Fund. Compensation insurance is one step among many social actions to assure conservation of human resources. The Fund led the way in writing insurance and provided for stabilization of an equit able rate structure; and it is needed to serve a segment of employers who cannot be appropriately served by other carriers. The organization is claim oriented and introduces integrity into the system through equitable and liberal action. The present day Fund has developed an expertise not available elsewhere on employer needs, provides a guarantee of equitable rates, and acts as an inhibitor of federalization which, many experts contend, would be less effective than a more localized system. California statutes governing workmen's compensation require the presence of the State Fund. 246 247 APPENDIX I (CONTINUED) ORIGIN: State Fund organizational design, calling for the provision of complete workmen's compensation service through field offices, is consistent with and fulfills the intent and spirit of the Boynton Act (1913). As the focal point of all service to our clientele, district office effectiveness in furnishing service is the critical measure of the State Fund performance as an insurer. BROAD GOALS: The goals of the South Pasadena District Office are: To optimize services and benefits to all California employees (both those injured on the job and those exposed to potential injury on the job) with the least stress upon the California society; and to concurrently provide State Fund employees the most personally rewarding involvement that is possible. This will (among other things) generate in the district (1) a positive image of the State Fund based on merit and earned respect, (2) an effective, responsive, sensitive, action-oriented agency, (3) equitable claims adjustment, and (4) fair and adequate premium charges. EXPECTED RESULTS: (1) Write a total of 43% more new business during the pilot year* than in the calendar year 1966. (2) Achieve a 15% net premium gain, using as a base period the calendar year 1966. (3) Create and maintain a pool of people who are characterized by a high degree of the following qualities: Self- awareness, growth, development, stress resistance, resili ence, independence, empathy, capable of love, having perception free of distortion, and feeling-tone based interpersonal relationships. All of these attributes are encompassed by the term "self-actualization." (4) Every candidate for CIR III will prepare himself to pass the written promotional examinations. (5) We will have at least the same percentage of candidates who successfully compete in the CIR IV examination as the Fund overall ratio. (6) Developmental reassignment will be initiated by the individual. (7) The perceived way in which the District Manager relates to his subordinates, and the perceived way in which supervisors relate to their subordinates shall show im provement to a statistically significant degree as measured on the Fleishman, Kirkhart and Gardner scale. *Pilot Year is July 1, 1967— July 1, 1968 248 APPENDIX I (CONTINUED) (8) All of the above to be accomplished through the utili zation of participative management. The management style assumes that the people selected and retained in the insurance classifications are positively charged, con structive, potentially self-actualizing individuals. The management climate will be one that fosters creativ ity, innovation, open communication and maximum partici pation in goal-setting and decision-making. (9) The perceived way in which the insured employer and his employees view the State Fund will show statistically significant improvement. (10) Preventable automobile accidents and lost time accidents will be less than the Fund average. (11) Information developed in the study should assist Fund efforts to improve internal relationships, develop higher employee morale and make additional provisions for em ployee and organization growth and development. PROGRAM STATEMENT PART II WORK PROGRAM: The work plan for achieving the expected results follows. This plan represents the thinking of South Pasadena District Office employees as of the date of its submission. Time consuming activities allocated were predominantly based on data available as of June 30, 1967. It is anticipated that the work program will change during the pilot year, the change being dictated by developing experience. Changes made in the work program will be recorded. The work program of the South Pasadena District Office is designed to involve all employees in the significant problem analysis and decision making activities of the district. This program is based upon an "open society" engaged in pursuing Fund, district, and individual goals in a manner that benefits the organization and the public. In the process of goal achievement the responsibility of adjusting to the environment is considered a district responsibility. Activities which district personnel indicate as probable approaches to achieving Expected Results are divided into (1) General Work Program, (2) Training Program, (3) District Office Performance Review Program, and (4) Additional Work Programs. 2 49 250 APPENDIX I (CONTINUED) 1. General Work Program A. New Prospect Evaluation and Solicitation Methods Identification of Prospect Prospect Processing Solicitation Regular prospects & regular po tential prospects. Key Accounts Bureau Card, when received, is checked to see if the firm is currently & active prospect. Prospects are screened for de sirability & then allocated among salesman. If there is a pro spect card, the bureau card is added to it. Or if not then the salesman makes a primary solici tation of the business (at this point name of carrier, exper ience modification, renewal date and spot evaluation of nature of account is made). Prospect card is fed into file by renewal date. Two months prior to re newal these cards are then redis tributed to salesman. Placed on key prospect list for con tinual soli citation through mon thly contact (letters tailored to the type of account and other comm unications) and at least quarterly personal calls. Preliminary direct mail solicitation including name of salesman who will call on the account. One call will be made prior to renewal. A follow-up will be made if re sponse is posi tive. Selective han dling which entails, as needed, a team approach which may include claims, safety legal, auditing and sales personnel. 251 APPENDIX I (CONTINUED) 1. A. Continued Individual Sales Commitments: Individual sales commitments will be made by each district representative which, when added up, amount to sufficient premium to achieve the expected result of a 43% increase. Emphasis will be placed upon the meaningfulness of the individual meeting his personal commit ment rather than upon highly competitive ranking by gross achievement. B. Servicing and Renewal Program: All accounts are considered to be new accounts irrespective of the length of time they have been with the Fund. In addition, renewal results are the collective res ponsibility of the personnel of the District Office. A part of this work program is to attempt to develop responsible individuals, not individual respon sibility, and to exert this responsibility collectiv ely. MONTHLY QUARTERLY EVENTS EVENTS ANNUAL EVENTS KEY Safety call— APSR delivery A) Year-end APSR ACCOUNTS Safety meet and presen presentation and ($25,00 & ing tation by discussion. over annual Inspection Safety and B) Dividend delivery premium and Schedule rate Field Repre conference. public inspection sentatives, This includes agencies) (where in plus Claims Sales, Safety, dicated) Representat and Claims Re Presentation ives, if presentatives . of special appropriate. C) TMO or modified training TMO where indi Programs cated. This (essentia could include lly) for Safety, Claims, supervisors. Sales, Legal, and One of these Auditing Repre activities sentatives. per month. APPENDIX I (CONTINUED) 1. B. Continued 252 QUARt'EKEY EVENTS MONTHLY EVENTS ANNUAL EVENTS KEY ACCOUNTS (cont'd) D) Annual audit, in cluding sugges tions for im provement in ac counting methods. E) Experience modi fication delivery presentation. F) Renewal call, in cluding intro duction of account team. G) Policy delivery call including introduction of account team. STANDARD See "Key” Same activities as ACCOUNTS accounts performed in con (§10,000 to activities junction with "Key §25,000) performed accounts. Some on a mon activities will be thly or eliminated due to quarterly lower premium size basis. or account sensi Selected activities from these columns will be performed. tivity. 253 APPENDIX I (CONTINUED) !• B. Continued 'SEMI - " ANNUAL MONTHLY EVENTS ANNUAL EVENTS SMALL ACCOUNTS ($3,000 to $10,000) Selected Selected activities activities from the list of from the list annual activities of quarterly on $10,000 to activities on $25,000 accounts. $10,000 to $25,000 accounts. Manpower allocation for the New Prospect Evaluation and Solicitation Methods Program and the Servicing and Renewal program.* New Prospect Evaluation and Solicitation Methods Program. Type of Account Key______Standard Small Function District Rep. Safety Rep. Auditing 23% 10 2.5 14% 5 2.5 7% 5 5 Total Hours Allocated = 9,297 * Taken together the percentages below add up to 300% which reflects 100% allocation of time for each of these functions (sales, safety and auditing). There are 27,104 manpower hours allocated to these two program areas out of the total of 45,120 hours available from the professional personnel. 254 APPENDIX I (CONTINUED) 1. B. Continued Type of Account Key Standard Small Function District Rep. 32% 18% 6% Safety Rep. 32.5 27.5 20 Auditing 53 20 17 Total Hours Allocated = 17,817 2. District Office Performance Review Program A. Budget Review. District personnel will review ex penditures and their effects. They will then re evaluate plans within the constraints of the dollar budget. B. Information System. rime will be expended in the development of a management reporting system appro priate to the program needs of the South Pasadena Office. Format of this system will be decided upon within the participative pattern in effect in the district. Continuing action will be taken to find more meaningful measures for effectivess of activities. 3. Training Program A. Developmental Training. Through collaboration train- ing goals common to the group will be explored and when feasible in-service training programs will be developed or contracted for. Individual felt- deficiency training and individual felt-growth training which cannot be met through the above will also be pursued and encouraged. Both of these will be designed to facilitate the growth of the personnel Total Manpower Hours Allocated = 11,280 APPENDIX 3. A. B. I ! , ) 255 I (CONTINUED) Continued and will be undertaken through collaboration and con currence of the participants. Complete control of training resides in the District. Outservices Laboratory Training (Based on total exposure of all employees and 50% re-exposure during pilot year). Reading Course Membership in Professional and Community Organizations Miscellaneous Training Expense NO. OF PEOPLE 24 29 29 23 Total Manpower expended on reading course: Total manpower expended in connection with specialized training, CIA general ized training and small group pro fessional meetings: Total TOTAL COST $ 2,160 4,350 5,220 1,725 2,000 $15,455 928 hr s. 5,900 hrs. 6,828 hrs. Accident Prevention. Employees will jointly review district safety achievement and will continue to provide regular and timely fleet inspections, defen sive driving training and district office inspection programs. 256 APPENDIX I (CONTINUED) 4. Additional Work Programs A. Underwriting. South Pasadena will continuously eval- uate and handle directly all accounts over $3,000. Underwriting will directly manage accounts under $3,000. The district will contract with home office staff for information and statistics on classifications and types of exposure, both statewide and Fund wide. Risks insured as a result of solicitation by South Pasadena personnel will be evaluated in the field and the underwriting department will process the appli cation. The South Pasadena Office will have the authority to issue all types of certificates of insurance. B. Claims Communications and Cost Control. On all cases involving: 1) Litigation 2) Obvious Permanent Disability 3) Temporary Disability over 90 days 4) Problem cases identified by the employer or SCIF The following items will immediately be provided to whichever person (highest possible level) the employer designates to receive them: 1) Medical Reports 2) Correspondence from the injured and a copy of our reply (1) 3) Awards and the Referee's notes 4) Recommended ratings 5) Compromise and Release approvals In addition to the sending of the above items, the employers representative would be contacted under the following circumstances: 1) Prior to any C & R negotiations. 2) Notice of WCAB hearings and our attorney's name, if known.(2) 1 Cf. Appendix II 2 Cf. Appendix III ! l 257 APPENDIX I (CONTINUED) 4. B. Continued In order that accounts over $25,000 be will informed on the status of all claims the Senior Adjuster will meet with the employer (highest possible level) once every three months to discuss any claims problems. This discussion should also involve the District Office Safety Representative and District Represen tative. C. Support Services. Support services supplied by the General Administrative Division, Program Development Division or other portions of the Fund will be charged to the South Pasadena District Office. 1) Business Services — All business services not retained or controlled by the Department of Gener al Services or the Department of Finance shall be the responsibility of the District Office. This includes the authorization to negotiate with other state agencies such as the Division of Purchases and to contract with State Fund Business Services. Some examples of the responsibilities of the district are: Purchasing Travel Advances and Expenses Automobile Procurement, assignment and utilization (including use of private automobiles) 2) Accounting Services — Budgeting and accounting services compatible with the uniform system of accounts and constraints of the Board and the State Controller will be performed by special arrangement with the accounting office. All other accounting functions will be performed by the district. 258 APPENDIX I (CONTINUED) 4. C. Continued 3) Personnel Services — Hiring, evaluating, classi fying, re-assignment, promoting, and placing employees within the district will be accomplished at the district level (and within the authority exercised by the Fund) unless the Personnel office performs such services by request and under contract. Personnel may be assigned to and from the South Pasadena District Office with the con currence of the district. 4) Legal Services — Legal services will be provided by a designated group of attorneys in the Los Angeles office who will provide a continuity of legal services in regard to filings against em ployers in the Los Angeles area. 5) Data Processing — As in the case of other central services, Pasadena will pay for services received. Services which exceed those extended to other offices will be subject to negotiation and additional charge. 6) Reports (The possibility of doing the following will be explored) a. Obtain information direct from the California Inspection Rating Bureau. b. Receive all statistics and information cir culated to Operating Divisions. c. Discontinue "range" figures from South Pasadena APSR's. d. Obtain from the actuary a dividend scale that provides for bigger payoff for major premium accounts and a dividend scale to be imposed on retrospective accounts. e. Obtain a monthly report of all outstanding audits categorized as 30, 60 and 90 days. D. Claims. A line relationship will be created between the District Claims Supervisor and the South Pasadena District Manager. Support services will be contracted for. For purposes of this experiment this relation ship should not be changed in a manner that results in a civil service change in the District Managers classification. 259 APPENDIX I (CONTINUED) 4. E. Communications with Home Office. The District Manager will report to the General Manager. The report will be in person or in writing as the situation dictates. SUMMARY A number of alternative methods are available to carry on the business of a Fund district office. The approach outlined in this program statement is a preliminary step to testing the "market" approach to managing such an office. Under such an arrangement the office becomes a "cost center" where both expen ditures and sales can be examined and a "profit" posture ascertained. The cost of the work program is deliberately held at the prev iously allocated level with the addition of estimated costs of a number of services that are normally included in general administrative costs. Giving visibility to such administrative costs should result in better and more thoughtful planning at both district and Fund levels. Costs are estimated as follows: South Pasadena District Office $ 400,000_______ "R" Zone, Los Angeles Claims $ 392,000_______ Services* $________________ Other Administrative Costs* $______ __ Total Budget $_ Recommended: Date___ District Manager Approved: _____________________ __ Date___ General Manager * We do not have information at this time which would enable us to estimate these costs. APPENDIX II 261 APPENDIX II (CONTINUED) NOTICE OF HEARING Hearing on ____________ Dept., is scheduled for ,_________ (Day) (Month) __, at _____________o'clock at WCAB )______ (Year) (Location) division office located at _____________________________ (Street Address) _________. The attorney representing you is (City) . You may contact him at Ext. APPENDIX III APPENDIX III (CONTINUED) Re: Injured's Name Date of Injury Fund Case Number In order to aid in your claims cost control we are enclosing the indicated information for your records: Medical Report Correspondence from Injured Recommended Rating Award Compromise & Release Other If you wish to discuss this case further, please feel free to contact us. APPENDIX IV 265 STATE 9 * 9 COLOCN 0 * » * A V fN Uf . SAN CAANCI«CO CAtl »f '*N IA • TCk. <4 t»> 9 0 1 - 1 0 * 1 C O V ,P K N i* A T iO N I N b U H A N C f FUND June 25, 1968 Dear Fellow Employees: PROGRAM ADVISORY COUNCIL As we continue to move toward shortening communication lines and placing more oopender.ee In district offices and support departments to achieve programmed objectives, we need your help. For this purpose, the Program Advisory Council Is being created. The purpose of the Council is to draw on the front-line experience, ideas, suggestions, and attitudes of the people who are out tnorc where the action Is. Specifically, It is hoped that the group can Identify and exchange; ideas on areas where programs need revision, clarification, simplification, better criteria or standards for measurement, or even deletion or. abanuonment. Initially, the Council will consist of two representatives selected by tr.e staff of each district office and one representative from Accounting, Data processing, Insurance Services, and Underwriting Departments. The selection (or election) of representatives rests entirely with each office or department, and is for a one-year term commencing July 1 each year. The Council will meet with representatives or the General Manager's Office at regular Intervals not less frequently than six times during each twelve-month term. Please send the names of those representatives selected to the attention of Alice Bowman, Secretary, by July 10, 1?66, in order that plans for an Initial meeting can be completed. Gener; M A H A P I M t l S O O * 1 0 7 . M 4 W C 6 C O 9 4 1 0 1 APPENDIX V STATE 0 f 4 OOLOCN G * ? l AVlNL't SAN rM ANCItCO CALIFOANIA • TCL. . < ? « • 0 O I I C ? » C O M P E N S A T ■ <.**?/ I N 8 U l( C I! FUND September 13, 1967 Dear Fellow Employees: ORGANIZATIONAL! RESEARCH PROJECT - MARKETING THEORY An experimental project is now under way to test the effect on District Offices of variations in organizational environment, delegation of authority, and management style. The project v;ae designed and will be conducted with the guidance of our consultant, Neely Gardner, and his assistant, Larry Klricnart. The test period is July 1, 1967 to July 1, 1906. These essential program elements are involved: South Pasadena District Office (District 'R'J'has been selected as tiie Pilot District. Other selected offices v.ill serve as test control points. Tne Pilot District is separated from the existing organisational structure. The District Manager, bill Rirconsin, will report direct to me on activities, programs, and conditions in the District. The District Manager will conduct the business of the Pilot District v . ’ ithin the same authority and the same constraints which apply to the General Manager unaer Corporate Policy, Board of Directors Policy, statutory anti constitutional provisions applicable to the State Fund. The test program of the Pilot District is geared to support the total programs of the State Fjnd. Management style will permit maximum participation by the District Office staff. The Pilot District will function as a profit center with responsibility for controlling operating costs within the financial limits provided by the Estimate of Expense arid the Program Statement. MAIL 0CX 007. CAN F A ANCI0C9. CALIFORNIA 04101 268 All existing services now provided to the Pilot District will be continued uriJ.oss specific changes are Initiated and arranged oy the District Manager. The approximate cost of all added or deleted services will be determined and cither charged or credited to the Pilot District. I am convinced that the test program can produce substantial advantages in the continuous effort to increase the effectiveness of our organisation and our satisfaction in our work. Your support of tula project is requested. GcnerasrManage' RAY s Jv APPENDIX VI 270 VARIABLE 1: RULES ORIENTATION E A UA Ti 17.4545 19.1364 18.7273 All Respondents T2 t3 19.6818 21.2273 17.8636 19.3182 20.2353 20.3182 T4 18.8182 20.4545 19.8421 E A UA T1 18.3750 19.1250 20.0000 Four-Time Respondents 18.5000 20.8750 20.1250 T3 17.8750 20.8750 19.8750 T 4 17.6250 19.5000 19.3750 VARIABLE 2: NURTURANCE OF SUBORDINATES E A UA Ti 23.4545 27.8182 26.4545 All Respondents t2 t3 26.5909 27.9091 27.0445 26.0909 27.7059 27.0000 T4 26.7273 27.1364 25.7895 E A UA Ti 24.0000 29.0000 27.0000 Four-Time Respondents t2 t3 28.2500 27.7500 27.2500 27.2500 27.3750 26.2500 T4 27.6250 27.0000 25.0000 GUIDE TO SYMBOLS AND TERMS: Scores = Arithmetic Mean All Respondents = 22 subjects Four-Time Respondents = 8 subjects E = Experimental Group A = Aware Control Group UA = Unaware Control Group T1 = Fall, 1967 T2 = Spring, 1968 T3 = Fall, 1968 T4 = Summer, 1969 271 VARIABLE 3: CLOSENESS OF SUPERVISION VARIABLE 4: UNIVERSALISM E A UA Ti 10.7727 14.4545 17.0455 All Respondents T2 t3 13.7273 15.1818 14.8182 13.2727 14.3324 15.9091 T4 12.9545 13.6818 15.3684 Four-Time Respondents E A UA T1 10.8750 16.5000 19.3750 12.2500 15.1250 17.1250 12.5000 14.7500 16.6250 11.7500 13.8750 16.2500 All Respondents E A UA Ti 16.2273 20.0909 20.3636 19.0909 20.4545 19.7727 19.3636 19.4706 19.5909 t4 18.8182 18.8182 19.4737 Ti Four-Time Respondents t2 t3 t4 E 17.7500 18.2500 19.2500 17.5000 A 19.1250 20.0000 19.5000 19.5000 UA 21.0000 20.7500 19.8750 19.3750 272 VARIABLE 5: PROMOTION- ACHIEVEMENT VARIABLE 6: PARTICIPATION IN DECISIONS All Respondents E A UA 15.7273 20.0000 18.5909 18.0455 19.1364 18.5000 18.0455 19.4706 19.0000 17.9091 19.8636 18.2632 E A UA Tl 15.7500 20.0000 19.6250 Four-Time Respondents T2 t3 19.0000 19.1250 18.3750 19.2500 19.3750 19.0000 T4 19.1250 18.5000 19.1250 E A UA Ti 22.0455 22.1364 22.9091 All Respondents T2 t3 21.6818 23.7727 22.0909 22.0000 22.6471 23.3636 T4 22.2727 21.8636 22.8421 Ti Four-Time Respondents t2 t3 t4 E 22.1250 21.6250 24.5000 26.2500 A 21.8750 22.5000 21.2500 25.0000 UA 23.0000 20.2500 20.8750 26.6250 273 VARIABLE 7: JOB TENSION VARIABLE 8: WORK-RELATED CHANGE E A UA Ti 30.1364 31.8636 29.7273 All Respondents T2 t3 31.0909 30.9091 30.8182 32.2273 30.6471 29.1818 T 4 31.7727 29.2727 30.8421 E A UA Ti 28.6250 32.6250 26.6250 Four-Time Respondents t2 t3 33.1250 34.5000 29.2500 32.6250 32.7500 27.1250 T4 30.1250 32.0000 32.0000 All Respondents E A UA T1 28.7273 30.4091 29.0909 29.0000 29.4545 29.4091 29.7727 28.7059 28.9545 t4 30.7273 26.5909 30.0000 T1 Four-Time Respondents E 29.5000 30.6250 28.0000 31.2500 A 30.2500 28.6250 28.6250 27.7500 UA 30.2500 30.3750 29.3750 29.5000 VARIABLE 9: SELF-ACCEPTANCE E A UA E A UA VARIABLE 10: ACCEPTANCE OF OTHERS E A UA 274 All Respondents *1 136.7727 133.5000 140.5000 t2 130.7727 136.4545 137.4091 t3 127.6364 136.2941 139.4091 t4 136.2273 121.8636 132.8421 Four-Time Respondents Ti 136.6250 127.1250 148.1250 t2 138.7500 131.3750 139.5000 T3 123.2500 132.5000 148.0000 T4 142.2500 121.3750 136.5000 All Respondents t2 t3 T1 94.4091 98.0455 101.4545 97.3182 98.0909 100.0000 96.2727 95.7647 100.8636 t4 99.5909 94.8636 98.0526 Four-Time Respondents T1 T2 T3 T4 E 98.7500 101.0000 98.7500 98.6250 A 95.8750 96.1250 97.2500 89.8750 UA 104.5000 106.2500 105.7500 97.0000 SELECTED BIBLIOGRAPHY BOOKS Allport, Gordon. Becoming. New Haven, Conn.: Yale University Press, 1955. Arygris, Chris. Intervention Theory and Method. Reading, Mass.: Addison-Wesley Publishing Company, 1970. ________. Organization Development. New Haven, Conn.: Yale University Press, I960. Beckhard, Richard. Organization Development: Strategies and Models. Reading, Mass.: Addison-Wesley Publishing Company, 1969. Bennis, Warren. Changing Organizations. New York: McGraw-Hill Book Company, 1966. ________. Organization Development. Reading, Mass.: Addison- Wesley Publishing Company, 1969. Bennis, Warren; Benne, Kenneth; and Chin, Robert, eds. The Planning of Change. New York: Holt, Rinehart an5 Winston, 1961. Berger, Peter, and Luckman, Thomas. The Social Construction of Reality. Garden City, N. Y.: Doubleday & Company, Inc., 1966. Blake, Robert, and Mouton, Jane. Building a Dynamic Corporation Through Grid Organization Development. Houston, Texas: Gulf Publishing Company, 1968. ________. Corporate Excellence Through Grid Organization Development. Houston, Texas: Gulf Publishing Company, 1968. ________. The Managerial Grid. Houston, Texas: Gulf Publishing Company, 1964. Blake, Robert; Shepard, Herbert; and Mouton, Jane. Managing Intergrouip Conflict in Industry. Houston, Texas; Gulf Publishing Company, 1964. 275 Boulding, Kenneth. The Image. Ann Arbor, Mich.: University of Michigan Press, 1961. Bradford, Leland T.: Gibb, Jack R.; and Benne, Kenneth, eds. T-Group Theory and the Laboratory Method. New York: John Wxley & Sons, Inc., 1964. Bruyn, Severyn. The Human Perspective in Sociology. Englewood Cliffs, New York: Prentice-Hall, Inc., 1966. a Chandler, Alfred, Jr. Strategy and Structures. Boston: M.I.T. Press, 1962. Dreitzel, Hans Peter, ed. Recent Sociology. Vol. II. New York: McMillan Company, 1970. Ellul, Jacques. The Technological Society. New York: Alfred A. Knopf, Inc., 1964. Etzioni, Amitai. The Active Society. New York: Free Press,1968 Farrell, R. Barry. Approaches to Comparative and International Politics. Evanston, 111.: Northwestern University Press 1966. Fleishman, E. A.; Harris, E. T.; and Burtt, E. H. Leadership and Supervision in Industry. Columbus, Ohio: Ohio State University, 1955. Fordyce, Jack, and Weil, Raymond. Managing With People. Reading, Mass.: Addison-Wesley Publishing Company, 1971 Fuller, R. Buckminster. Operating Manual for Spaceship Earth. Carbondale, 111.: Southern Illinois University Press, 1969. Galbraith, John Kenneth. The New Industrial State. Boston: Houghton Mifflin Company, 1967. Gordon, Chad, and Gergen, Kenneth J. The Self and Social Interaction. New York: John Wiley & Sons, Inc., 1968. Heady, Ferrel, and Stokes, Sybil. Papers in Comparative Administration. Ann Arbor, Michigan: University of Michigan Press, 1962. Janov, Robert. The Primal Scream. New York: G. P. Putnam's Sons, 1970. Kahn, Robert; Wolfe, Donald; Quinn, Robert P.; and Snoek, J. Diedrick. Organizational Stress. New York: John Wiley & Sons, Inc., 1964. 277 Katz, Daniel, and Kahn, Robert. The Social Psychology of Organizations. New York: John Wiley & Sons, Inc., 1966. Laing, R. D. The Divided Self. London: Tavistock Publications, 1960. Lawrence, Paul, and Lorsch, Jay. Organization and Environment. Cambridge, Mass.: Harvard University Press, 1967. Lippett, R.; Watson, J.; and Westley, B. The Dynamics of Planned Change. New York: Harcourt, Brace & World, Inc., 1961. Lowen, Alexander. Pleasure. New York: Coward-McCann, Inc., 1970. Mannheim, Karl. Man and Society in an Age of Reconstruction. New York: Harcourt, Brace & Co., 1946. March, James, ed. Handbook of Organizations. Chicago: Rand McNally & Company, 1965. Marini, Frank, ed. Toward a New Public Administration. Scranton, Pa.: Chandler Press, 1971. Marrow, Alfred; Bowers, David G.; and Seashore, Stanley E., eds. Management by Participation. New York: Harper & Row, Publishers, 1967. Peris, Frederick. Gestalt Therapy Verbatim. Palo Alto, Calif.: Real People Press, 1970. Presthus, Robert. The Organizational Society. New York: Random House, Inc., 1$65. Raskin, Marcus G. Being and Doing. New York: Random House, Inc., 1971. Rogers, Carl. On Becoming a Person. New York: Houghton Mifflin Company, 1959. ________. On Encounter Groups. New York: Harper & Row, Publishers, 1970. Rosenau, James. N. Linkage Politics. New York: Free Press,1969. Roszak, Theodore. The Making of a Counter Culture. Garden City, N.Y.: Doubleday & Company, Inc., 1969. Schein, Edgar. Process Consultation. Reading, Mass.: Addison- Wesley Publishing Company, 1969. 278 Schutz, Alfred. The Phenomenology of the Social World. Evanston, . 111.: Northwestern University Press, 1967. Stanley, Julian, and Campbell, Donald. Experimental and Quasi- Experimental Designs for Research. Chicago: Rand McNally & Company, 1966. Taguri, Renato, and Litwin, George, eds. Organizational Climate. Cambridge, Mass.: Harvard University Press, 1968. Toffler, Alvin. Future Shock. New York: Random House, Inc., 1970. Touraine, Alain. The Post-Industrial Society. New York: Random House, Inc., 1971. Veldman, Donald J. Fortran Programming for the Behavioral Sciences. New York: Holt, Rinehart & Winston, 1967. Vroom, Victor. Some Personality Determinants of the Effects of Participation. Englewood Cliffs, N. J.: Prentice- Hall, Inc., i960. Waldo, Dwight. The Administrative State. New York: Roland Press, 1947. Zyderveld, Anton C. The Abstract Society. Garden City, N. Y.: Doubleday & Company, Inc., 1970. ARTICLES FROM BOOKS Benne, Kenneth. "History of the T-Group in the Laboratory Setting." T-Group Theory and the Laboratory Method. Edited by Leland T. Bradford; Jack R. Gibbs; and Kenneth Benne. New York: John Wiley & Sons, Inc., 1964. Biller, Robert. "Adaption Capacity and Organizational Develop ment." Towards a New Public Administration. Edited by Frank Marini. Scranton, Pa.: Chandler Press, 1971. Bowers, David G., and Seashore, Stanley E. "Employee Attitudes, Motivations and Satisfactions." Management by Partici pation. Edited by Alfred J. Marrow; David G. Bowers; and Stanley E. Seashore. New York: Harper & Row, Publishers, 1967. 279 Chin, Robert. "The Utility of Systems Models and Development Models for Practitioners." The Planning of Change. Edited by Warren Bennis; Kenneth Benne; and Robert Chin. 2nd ed. New York: Holt, Rinehart and Winston, 1961. Dorsey, John. T., Jr. "An Information-Energy Model." Papers in Comparative Administration. Edited by Ferrel Heady and Sybil Stokes. Ann Arbor, Mich.: University of Michigan Press, 1962. Greiner, Larry E.; Leitch, D. Paul; and Barnes, Louis B. "The Simple Complexity of Organizational Climate in a Governmental Agency." Organizational Climate. Edited by Renato Taguri and George Litwin. Cambridge, Mass.: Harvard University Press, 1968. Kirkhart, Larry. "Public Administration and Selected Develop ments in the Social Sciences." Toward a New Public Administration. Edited by Frank Marini. Scranton, Pa.: Chandler Press, 1971. Marrow, Alfred. "Planning the Changes." Management by Participation. Edited by Alfred J. Marrow; David G. Bowers; and Stanley E. Seashore. New York: Harper & Row, Publishers, 1967. Rosenau, James N. "Pre-Theories and Theories of Foreign Policy." Approaches to Comparative and International Politics. Edited by R. Barry Farrell. Evanston, 111.: Northwestern University Press, 1966. Schroyer, Trent. "Toward a Critical Theory for Advanced Industrial Society." Recent Sociology. Vol II. Edited by Hans Peter Dreitzel. New York: Macmillan Company, 1970. Scott, W. Richard. "Field Methods in the Study of Organization." Handbook of Organizations. Edited by James March. Chicago: Rand McNally & Company, 1965. Shepard, Herbert. "Changing Interpersonal and Group Processes." Handbook of Organizations. Edited by James March. Chicago: Rand McNally & Company, 1965. Tiryakian, Edward. "The Existential Self and the Person." The Self and Social Interaction. Edited by Chad Gordon and Kenneth J. Gergen. New York: John Wiley & Sons, Inc., 1968. ARTICLES IN JOURNALS Bennis, Warren. "A Funny Thing Happened on the Way to the Future." American Psychologist, XXV (October, 19701, 391-410. Berger, Emmanuel. "The Relation Between Expressed Acceptance of Self and Expressed Acceptance of Others." Journal of Abnormal and Social Psychology, XLVII (April, 1952), 778-82. Blake, Robert. "A Second Breakthrough in Organization Develop ment." California Management Review, XI (Winter, 1968), 73-78. Blake, Robert; Mouton, Jane; Barnes, Louis; and Greiner, Larry. "Breakthrough in Organization Development." Harvard Business Review, XL (November/December. 1964), 133-55. Burke, W. Warner. "Management Development and Organization Development." Journal of Applied Behavioral Science, VII (September/October, 1971), 569-79. Campbell, Donald. "Factors Relevant to the Validity in Social Settings." Psychological Bulletin, LIV (April, 1957), 297-300. "Reforms as Experiments." American Psychologist, XXIV (April, 1964), 411-21. Crockett, William. "Team Building— One Approach to Organization Development." Journal of Applied Behavioral Science, VI (July/August/September, 1970), 291-306. 4 Davis, Sheldon. "An Organic Problem - Solving Method of Organizational Change." Journal of Applied Behavioral Science, III (January, 1967), 3-21. Emery, F. E. "The Next Thirty Years." Human Relations, XX (Fall, 1967), 199-237. Emery, E. F., and Trist, E. L. "The Causal Texture of Organizational Environments." Human Relations, XVII (February, 1965), 21-37. Friedlander, Frank. "Behavioral Research as a Transactional Process." Human Organization, XXVII (Winter, 1968), 369-379. 281 Hampden-Turner, C. M. "An Existential Learning Theory and Integration of T-Group Research." Journal of Applied Behavioral Science, II (October/November/December, 1966) , 367-386. Hauser, Philip M. "The Chaotic Society." American Sociological Review, XXXIV (September, 1968), 1-19. Katz, Daniel. "Group Processes and Social Integration." Journal of Social Issues, XXXIII (June, 1967), 3-22. Seashore, Stanley E., and Bowers, David G. "Durability of Organizational Change." American Psychologist, XXV (April, 1970), 227-33. Shepard, Herbert. "Innovation-Resisting and Innovation- Producing Organizations." Journal of Business, XL (June, 1967), 267-81. ________. "Personal Growth Laboratories." Journal of Applied Behavioral Science, VI (November, 1970) , 259-66. Terreberry, Shirley. "The Evaluation of Organizational Environments." Administrative Science Quarterly, XII (March, 1968), 590-613. Tolman, Edward C., and Brunswik, Egon. "The Organism and the Causal Texture of the Environment." Psychological Review, XLII (January, 1935), 43-77. Trumbo, Don. "Individual and Group Correlates of Attitudes Toward Work-Related Change." Journal of Applied Psychology, XLV (August, 196lJ^ 338-44. Wilson, Woodrow. "The Study of Administration." Political Science Quarterly, II (June, 1887), 197-222. 282 UNPUBLISHED MATERIALS Gardner, Neely. "Organizing for the Future— A Trial." Paper Presented at the 1971 National Conference of the American Society for Public Administration, Los Angeles, Calif., University of Southern California School of Public Administration, May, 1971. (Mimeographed.) Greiner, Larry. "Evaluating a Pilot Program Using Management Grid for Organization Development in the U. S. Forest Service." Unpublished pamphlet, U. S. Forest Service, n.d. Herman, Stanley M. "What is This Thing Called 'Organization Development'?" Unpublished paper, TRW Systems, Redondo Beach, Calif., approx., 1970. (Mimeographed.) Sherwood, Frank. "Devolution as a Problem of Organization Theory." Unpublished paper, University of Southern California School of Public Administration, n.d. (Mimeographed.) Trist, E. L. "The Relation of Welfare and Developments in the Transition to Post Industrialism." Paper presented at the Canadian Center for Community Studies, Attawa Seminar, November, 1967.

Asset Metadata

Creator Kirkhart, Larry Jack (author)

Core Title Organization Development In A Public Agency: The Strategy Of Organizational Devolution

Contributor Digitized by ProQuest (provenance)

School School of Public Administration

Degree Doctor of Philosophy

Degree Program Public Administration

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Political Science, public administration

Format dissertations (aat)

Language English

Advisor Storm, William Bruce (committee chair), Gardner, Neely (committee member), Newman, J. Robert (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-796924

Unique identifier UC11363541

Identifier 7314419.pdf (filename),usctheses-c18-796924 (legacy record id)

Legacy Identifier 7314419

Dmrecord 796924

Document Type Dissertation

Format dissertations (aat)

Rights Kirkhart, Larry Jack

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA

Linked assets

University of Southern California Dissertations and Theses