The ECCO System*:
Foundations for Total Quality Management
*Cybernetic Principles for Effective Control in Complex Organizations.
By Kevin B Kreitman
B.A., San Jose State University, 1978
M.S., San Jose State University, 1986
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Advanced Technology in the Graduate School of the State University of New York at Binghamton, 1992
Dissertation Committee: Dr. Robert Emerson; Dr. George Klir; Dr. Eileen Way
Copyright 1992 by Kevin B Kreitman
TQM (Total Quality Management) is a technology for achieving the transformation of organizations from their existing forms into a form which is capable of dealing with modern requirements. This modern form shall be referred to as an ECCO system: A system for Effective Control of Complex Organizations. In the U.S., industry is moving primarily from a bureaucratic Weber/Taylor form of organization to an ECCO form. The Weber/Taylor organizational structure and culture have trappings which are systematically hostile to ECCO system operations. Thus, when the tools and techniques associated with Total Quality are adopted without changing the fundamental infrastructure, Total Quality results will not be achieved.
The changes required to obtain a Total Quality organization, and to make effective associated techniques, practices and advanced manufacturing technology, are all based on certain cybernetic principles for the control of complex systems. The phenomenon of "control" in such systems is a combination of regulation and co-ordination. The role of a "controller" function is therefore twofold: Not only must it set the parameters of the target function to be regulated, it must also architect the interactions among system elements so that the system will tend to behave as desired, moving in a coordinated way toward the identified goal. The majority of management attention, and the bulk of TQM activities, often focus on the regulatory aspect of control, ignoring the critical role of co-ordination.
Nine ECCO system principles are identified as critical to successful Total Quality efforts. The first four are necessary to the regulatory activities of the organization, including: the closing and shortening of all feedback loops; maximization of the capacity of all first-line regulators; planning and problem-solving to be performed by a heterarchy of first-line regulators who, in effect, model the process under consideration; and the analysis and correction of process, not just inputs.
The next five principles are seen as critical to management and co-ordination efforts of the organization, including: the intrinsic alignment of incentives, evaluation systems, and reward structures for employees; of measurement systems; of company goals and all subsystem goals; of communications systems via common language and shared evaluation protocols; and finally, of oversight inspections or reviews which ensure that common meta-structure and meta-processes are being used and that goals are being achieved.
After providing the cybernetic foundations for the principles, a number of associated operational and management practices will be identified, as evidence for use of the principles. Finally, it will be shown that these practices are present in three independently recognized quality organizations, recipients of the Malcolm Baldrige National Quality Award.
To Mai Von Foerster with love and enormous admiration.
I had a specific agenda in choosing both the topic and my approach to research for this dissertation. Elucidating my personal interests in this work may clarify both its intended scope and its potential significance. I have had a long-time fascination with a seemingly endemic phenomenon: why things don't work the way they are supposed to, despite our apparent best efforts. For example: Why have a variety of the government policies related to welfare and "workfare" (such as CETA and Job Training Partnership Act) failed to make a dent in poverty, or improve the employability of the hard-core unemployed for whom the programs were designed? Why did deregulation of the transportation industry fail to increase competition, improve service, and reduce airfares and shipping charges as anticipated? Why is the stated intent of many educational institutions to provide high-quality instruction while the result is often mediocrity and frustration for students, teachers and administrators alike? Why do so many business enterprises aim for success, yet flounder and fail despite the apparent best efforts of all involved?
There are always specific reasons suggested for these failures. The poor don't want to succeed, employers discriminate, transportation companies are greedy or poorly managed, the teachers' union behaves unprofessionally, students care only about grades, administrators stifle academic freedom, finance managers won't invest in innovative designs, labor unions interfere with productivity, engineers are "prima donnas" and don't understand business. All of these explanations may be true in their various applications, yet they do not point the way to solutions. In fact, so stated, they divert attention from real problems and real solutions.
One of the fundamental observations in cybernetics1,2, underscored by recent work in chaos theory3, is that systems characterized by circular causal relationships have extremely rich and "unpredictable" behaviors, though they are essentially determinate. All systems with feedback, mutual or recurrent interactions are potentially included: ecosystems, human activity systems, biological systems, and ongoing human organizations and cultures. The behavior of these systems, and therefore their output, is a direct result of the system dynamics. Results are determined by what interactions actually take place, intended or not.
This fact has been demonstrated repeatedly by System Dynamics modelers over the past 30 years, in such diverse applications as ecosystems4, business cycles5, and public policy6. Some classic studies in the dynamics of research and development activities discovered hidden dynamics which inherently drove cost and time overruns, in the face of all efforts to prevent them7. The lesson here is that you cannot control a system (whose behavior is the result of circular dynamics rather than statistical randomness) until you understand the interactions, which create the behaviors. Seemingly sound policies fail to have the desired effects, and the reasons why remain a mystery--until the actual system dynamics are modeled. Control is not simply a result of intention.
In fact, intentions by themselves do not seem to matter a great deal. A system which is "mandated for excellence, but engineered for mediocrity" will almost inevitably produce mediocrity. Brief exceptions to this fact can be observed in human activity systems, fueled by superhuman efforts of people resisting the dynamics, rising above and beyond the call of duty until they burn out or are overwhelmed by the inertia of the system itself.
Thus, I was led into the study of cybernetics to find out why things (mis)behave as they do, and to discover how to diagnose and restructure them so that they tend to behave the way we wish them to. In looking at manufacturing organizations, I was impressed by the success of the Japanese, and by the apparent magic of Total Quality Management as espoused by W.E. Deming and Japanese authors Kaoru Ishikawa and Hajime Karatsu. But as I observed and later participated in such efforts, I realized that some fundamental understandings, some essential elements were often missing in the application. Companies would send their Quality Assurance people to learn Statistical Process Control, and use workers only to keep data charts. Process Action Teams and Quality Circles were instituted within departments, often with little training, no corporate direction except to "fix problems", and with inadequate training and facilitation. I watched workers be "empowered" by management, with inadequate training, direction and supervision, with sometimes disastrous results for the company. Sometimes "empowerment" simply allowed management to blame people lower down the chain of command for errors and problems. In short, many organizations have adopted bits and pieces of Total Quality strategies--sometimes lots of bits and pieces, at considerable cost--but have achieved little except disappointment, disillusionment and cynicism among workers and managers alike. What was missing did not seem to be specific practices so much as the way the practices were utilized and coordinated. This, too, seemed the source of the failure of organizations to adopt some complementary practices, such as adequate and appropriate training, or changes in management or departmental relationships. They simply could not see the relevance of that particular, isolated piece to their operations.
It is frustrating not to be able to communicate the essentials of what works and why it works, particularly when speaking to people who might potentially hear it. It is also frustrating to know that there is "something" which works, but to lack the clarity to identify it and communicate it. I was intrigued: what were the essentials, the underlying dynamic and coordinating structures which produced success, or some essential conditions for it? And how could an organization which was not successful reorganize itself and adopt practices to change itself into a successful organization?
One issue was clear. The old organizational structure which had heretofore provided control, the Weber/Taylor bureaucracy, was standing in the way of success. Yet the only things suggested to replace its function were "leadership" and "trust", hardly a recipe for a replacement structure or mechanism. Social and biological analogies (the organization as "clan"8 or as "organism") have been provided, but analogies tend to function better after the fact, or to give a feel for alternative ways of perceiving situations. It takes a lot of work and specific knowledge both of the source of the analogy and of the organization in question to make these analogies useful in terms of providing particular structures and practices which actually fix problems and improve performance.
I was looking for principles which would be at once more general and more specific than analogies or the current models offered: more general in that they would apply to a wide variety of organizations, not just manufacturing or even for-profit ones; and more specific in that they would give definitive guidelines and criteria for particular practices and structures necessary for successful control. After a review of the literature about quality in manufacturing organizations, I developed a feel for some principles I believed essential, based on my study of cybernetics and systems. I had trouble finding succinct models and evidence in the general systems literature, until I looked again at the work of William Ross Ashby. I had felt certain that some of the principles I sought would be found there. To my relief and satisfaction, virtually all could be found or developed from Ashby's formal work.
The particular approach I have used in developing the theoretical structures and principles has its strength in its formal generality: its application to a wide variety of systems and types of system components, from humans to technical practices to fully automated systems. It also has the virtue of being specific in terms of the criteria and purposes of various configurations of control mechanisms, and their necessary capabilities. These can be translated directly into mechanical, informational, and human capacities and requirements. Though the principles I have developed are, I believe, broad and suggestive of a particular infrastructure, I have not fully developed the infrastructure requirements here. More research is certainly necessary to do so, and I suspect it will be fruitful. However, the specificity of the principles should enable diagnosis and intervention in systems which are at present "mis-behaving" in spite of Total Quality efforts.
I firmly believe it is possible to have a successful Total Quality effort without understanding these formal foundations. In fact, I have seen several companies do so. However, this situation is like that of an individual who is able to be healthy, with low cholesterol, normal blood sugar, body weight and blood pressure without understanding the entwined physiological principles which maintain these parameters. An individual who has "good" heredity will be fine under a variety of inputs and circumstances. One who simply follows a physician's advice and directions exactly will probably be able to keep the parameters within bounds. But the understanding of underlying physiological principles should give one leverage: to improve one's health in the face of poor initial physical condition; to control and maintain a failing or stressed system; and to increase one's ability to control these factors under a wide variety of circumstances and stresses, without additional specific medical advice or medication. Of course, success in health, like success in Total Quality, has two major prerequisites: That one truly cares, and that the efforts be carried out with integrity. No theory, no knowledge, no formula will supplant the need for these.
I offer the following work as the foundations of a control infrastructure necessary for the success of modern manufacturing organizations. I believe that the principles will be found to have application to other organizations which find themselves constrained by complexity as well. I designate these principles as ECCO system requirements: the foundations for the Effective Control of Complex Organization. I contribute this work to stand in addition to the existing and invaluable management wisdom, innovative practices, and modern supporting technology for the ultimate success of Total Quality efforts in all organizations. I am also pleased to bring the brilliant seminal work of William Ross Ashby back into the mainstream of applied cybernetics, and offer it as an introduction for general readers that they might consider the strength of the paradigm and the insights it may generate.
There are many people I wish to acknowledge both for their instrumental assistance in the creation and production of this work, and for the more global influence they have had on my life to bring me to the point where I could produce it. These people include: Beatrice Cossey, who served as a powerful role model and was perhaps the sole influence strong enough to cause me to finish my bachelor's degree; my friend and colleague Bruce McKellar, fellow seeker of the truth (a rare commodity in academia); Richard Blum of Stanford University, who gave me my start in research when I was an undergraduate dropout, and who had seemingly unlimited faith that I would go on; to my parents, and my aunt and uncle who helped to make that possible; Ruth Ellen Miller, former director of the Cybernetic Systems Program at San Jose State University; Serena Stanford, AAVP for Graduate Studies and Research, and the other faculty and administrators at San Jose who encouraged me to continue on the road to my Ph.D.
I owe more than I can say to a few people: to my best friend and colleague Elin Whitney-Smith, for more reasons than I can possibly list, and who knows all of them; to Heinz and Mai von Foerster, particularly for their insights, love and encouragement; to my friend and colleague Christoph Berendes for friendship and critical review; to my friend and colleague Kirstie Bellman, for intellectual and material support; to my step-daughter Lilian for forbearance during the very trying year it took to produce this work; to my former husband, Thomas McCombs, who served as an inspiration and intellectual colleague; and finally to my parents, Rainy Lavin Kreitman and Seymour Keith Kreitman for raising me to believe in myself and my own judgement, to understand that being in the minority did not mean being wrong, and instilling in me the belief that there was nothing I could not do. And for the improvements and revisions being undertaken in the emerging version of this document, I owe a debt of love and gratitude to my beloved husband, Thaddeus Cooper, for love, encouragement, inspiration, constructive critique, and invaluable technical support.
Cybernetic Principles for Effective Control of Complex Organizations
Table of Contents
Table of Figures
3: Table of Outcomes for R and D (perfect regulation possible)
4: Table of Outcomes for R and D (perfect regulation not possible)
6: Sequential Regulation (without hierarchy)