|
July 16, 1995
Siam News W. Edwards Deming, the modest but outspoken man who died of cancer on December 20, 1993, was one of the world's best-known advocates for quality. He bestrode the management consulting world with a deep conviction that the solutions to our problems would be simple it we would only turn from the comfortable but ineffective attitudes of the past and embrace the concept of minimization of variation (along with all its very unfamiliar ramifications). Deming was born 0ctober 14, 1900, in Sioux Falls, Iowa, and grew up in Powell, Wyoming. Much has been made of the influence of his frontier upbringing on his beliefs in the importance of people, the value of cooperation, and the deadliness of waste. Whatever their source, there is no doubt that these values shaped his life and work and caused him to be the bearer of a message that many of the world's religious leaders, philosophers, and advocates of positive thinking would find familiar. Do your best, continually seek to improve that best, look out for the people you are responsible for, and recognize that everyone is in this together. The power of Deming's philosophy lies in his emphasis on how to do this, using statistical methods. In fact, Deming's influence on the statistical community in the 1930s and 1940s was considerable, but his greatest influence over the world at large came later in his life, when he branched out from statistics proper to put his energies into the use of statistical methods for the benefit of humankind. Deming's formal education included an engineering degree from the University of Wyoming, a master's degree in mathematics and physics from the University of Colorado, and a PhD in mathematical physics from Yale University in 1928. A Natural Progression of Interests On receiving his PhD, Deming was offered a job at the Western Electric Company, where he had worked summers, but chose instead to enter government service. The first of what would prove to be three careers, which lasted until 1939, was in the Fixed Nitrogen Research Laboratory of the U.S. Department of Agriculture. There, the study of the physical properties of materials led him into statistics via the theories of measurement errors and least squares that he encountered in this work. This was, he asserted, a "natural progression of interests." His path was also influenced by surveying and geodesy, where least squares methods were also used, and the kinetic theory of gases, from which he learned a lot of probability. Deming did not contribute to the foundational issues in probability that were occupying Kolmogorov at the time, but rather was drawn to applications of statistics. The probability notions of the day were not fully developed to our contemporary standards of mathematical rigor, but they were at least usable for those applications. It was during this period that Deming's statistical style developed, under the influence of Karl Pearson, Jerzy Neyman, Harold Hotelling, Sir Ronald Fisher, and Walter Shewhart. From Neyman he acquired the enumerative, descriptive statistical viewpoint. From Fisher (with whom he studied in London during a year's leave of absence from USDA) he acquired the predictive, inferential, analytical approach that he found of primary importance. Deming equated prediction with the providing of answers to the question "What would be best?" In other words, the statistician's task is to design the experiment (when possible), examine the data, and provide guidance on how to proceed in some human endeavor (an early expression of the decision-theoretic point of view). Deming felt that Pearson and Neyman had fallen into the trap of using enumerative methods on what were fundamentally prediction problems. At first, considered the X2 theory, type 1 and type 2 errors, significance, and all the other machinery of sampling useless for prediction (see his article "On Probability as a Basis for Action," The American Statistician, Vol. 29, 1975, pages 146-152). Later on, Deming would synthesize the predictive and the enumerative points of view in his work on industrial quality control; the famous "red bead" experiment, in 1942, is an example of how sampling, process capability, and improvement ideas work together. Deming was introduced to Walter Shewhart of the Bell Telephone Laboratories by Kunsman, who was vice-president of the Fixed Nitrogen Research Laboratory. Shewhart was a critical influence on Deming during this period, in which Deming began to move in the direction of his second career--the application of statistical methods to industrial production and management. Shewhart's idea of common and special causes of variation led directly to Deming's theory of management. Deming saw that these ideas could be applied not only to manufacturing processes but also to the processes by which enterprises are led and managed. This key insight made possible his enormous influence on the economics of the industrialized world after 1950. Deming invited
Shewhart, as well as Fisher, Wishart, Cochran, and others, to
lecture
at USDA. During this time, Deming wrote, among others, Statistical
Adjustments of Data (1938). He arranged for the printing (1940)
of Facsimiles of Two Papers by Bayes With Commentaries (Deming's
commentary concerned Bayes's use of divergent series; the other
was by Bell Labs's E.C. Molina, famous for early applications
of
queueing theory in telephony). He also edited the series of lectures
delivered by Shewhart at USDA, Statistical Method from the Viewpoint
of Quality Control, into a book in 1939. One reason he learned
so
much from Shewhart, Deming remarked in a videotaped interview,
was that, while brilliant, Shewhart had an "uncanny ability to
make things difficult"; Deming thus spent a great deal of time
both figuring out Shewhart's ideas and devising ways to present
them with his own twist. Quality:
Vital Role In 1939, Deming moved to the Bureau of the Census as head mathematician and adviser on sampling, beginning his second career. From 1928 to 1942, he taught statistics courses at the USDA Graduate School. With the outbreak of World War II, he also taught statistics courses at Stanford in support of the War effort. Later in life, he would be discouraged by his experience with those courses: The engineers and statisticians who were trained there did very well, but he did not feel that the managers, who attended only half-day familiarization sessions in statistical techniques, learned enough to see the benefits of their efforts of their engineers and technical personnel. The good work of the technical personnel then gradually died away for lack of support. Deming concluded that quality can be improved only if top management is part of the solution and participates appropriately and actively in the quality program. In a videotape he made in 1981 for the American Statistical Association (Deming on Statistics in Industry; Implications for Statisticians and Their Statistical Education), Deming shares credit with J.M. Juran for the realization of the vital role of upper management in quality. The failure of quality control to catch hold in the United States after the war, because of the lack of involvement by upper management, caused him to avoid the same mistake when he took the ideas of minimization of variation and quality management to Japan (1947-50). During this period, he was working for MacArthur's Supreme Command, Allied Forces, Tokyo, as adviser in sampling techniques. Starting in August 1950, Deming lectured on his statistical and management methods to the leading industrial companies in Japan. Under the auspices of the Japanese Union of Scientists and Engineers (JUSE), he brought together the top managers of these companies before training the staff in his technical ideas. His influence was so strong that by December 1950, JUSE had instituted the Deming Prize, an annual prize for excellence in quality. (Only two American companies have ever won the JUSE Deming Prize: Florida Power & Light Company and AT&T Microelectronics Power Systems.) Deming saw that
the top management of the Japanese companies was ready to use his
ideas, and he expressed his confidence that they would come to dominate
world markets inside five years. The managers themselves did not
believe this and wondered at the basis for his confidence. As it
turned out, the only erroneous part of Deming's prophecy was the
time scale; the American and European economies are feeling the
consequences to this day and surely will continue to do so for many
more years. During this time, Deming published Some Theory of Sampling
(1950). Deming's third career began in 1946, when he hung out his shingle as a consultant in statistical surveys. For the next 47 years, he worked with the energy of two, showing just how broad the notion of statistical surveys could be. He developed his "14 Points," the key actions management must take to ensure quality, productivity, and success. He summarized the "Seven Deadly Diseases" that prevent improvement in quality. He synthesized the Deming Cycle (sometimes called the Shewhart Cycles) of Plan-Do-Check-Act for guiding quality projects (indeed, all projects). He joined the facility of the New York University Graduate School of Business in 1946 and taught there long past whatever mandatory retirement age may have existed (he was still teaching there only a few months before he died). He also taught at Columbia University and maintained a schedule of four-day short courses that would have exhausted a person half his age (he completed one in Los Angeles only 10 days before his death). In addition to his consulting work with the top management of many American companies, with notable success at Ford and Xerox, Deming consulted with public agencies and private concerns in Greece, India, Japan, Germany, Austria, the U.K., France, Mexico, China, Turkey, Brazil, and Argentina. His third career was the full flowering of his philosophy, which was centered on people and the dignity of work. He believed that people should have joy in their work, that the system within which they work should be designed to make this possible and to enable workers to reach their full potential to contribute to the enterprise, that the system is management's responsibility, that 85% of all quality problems are management problems (that is, symptoms of a malfunctioning system), and that organizations and their suppliers need to work together to optimize results for both. (To illustrate the win-win outcome that he advocated for customers and suppliers, he sometimes used as an example the Great Western Sugar Company's practice in the 1920s of educating farmers--its suppliers--in the cultivation of beets.) Deming's statistical ideas had finally reached the point where he saw they would do the most good: in advancing the welfare of the country and society. Broader management
issues, rather than specific statistical topics, were the focus
of his publications in this period: Statistical Design in Business
Research (1960), Quality, Productivity, and Competitive Position
(1982), Out of the Crisis (1986), and The New Economics (1993). Deming Stories An interesting facet of Deming's influence is the proliferation of so-called Deming storiesCanecdotes in which he is seen putting his own ideas into practice. Peter Kolesar of Columbia University relates a Deming story that illustrates his approach to teaching and students. In the regular course of events during a semester at Columbia. Kolesar says, Deming was asked to have his students fill out course evaluation forms of the type well known to all academics. Deming was opposed to the idea on the grounds that his students, being students, were simply too inexperienced to identify what was truly valuable and that their evaluations would be in a deep sense meaningless. One can easily imagine him asking, in his gravelly voice, "How…can…they…know?" In fact, Deming said he would be interested in their opinions not at that time, but ten years later, when they had had a chance to mature and had been exposed to more of life (and to lots of examples of untamed variation causing undesirable results). Many such "Deming stories" are current. The power of these stories, and the reason they are passed eagerly from person to person, is that they illustrate Deming's willingness to speak truths that we all recognize, deep down, as real--and that we often fervently wish were acted upon--but that we won't speak because we are chagrined, possibly even ashamed, that while we believe them, we don't often act on them. One of the interesting foibles of our society is that we seem to be willing to let the elderly say these things--perhaps on the theory that what they say doesn't matter much, which, in Deming's case, would of course be a serious misjudgment. Some found him impatient, even brusque, but these were usually executives who were unaccustomed to being spoken to directly about their faults. Deming was in reality almost unbelievably patient with those who were eager to learn. After all, he had been answering the same questions in his courses over and over for 40 years. He has been called "the patron saint of quality." Some, though, think that Deming's program cannot succeed precisely because it is so utopian, requiring as it does actions guided by such easily damaged values as trust and teamwork, or because it is so hard for mere mortals to live up to the standards Deming set. But his program has had notable success: Ford Motor Company and Xerox Corporation in the U.S. and Toyota Motor Company in Japan are only a few of the most outstanding examples. Deming merely asks that we rise above ourselves, and he gives us a framework in which this can be a positive feedback system. Impossibly utopian? Not to everyone. Deming published over 170 books and papers. He was a pianist and composer: He wrote a version of "The Star-Spangled Banner" that was designed for ordinary people to sing. He also composed a number of Masses, including the Missa Spiritui Sancto for organ and chorus. He was awarded honorary degrees by 14 universities, the Second Order Medal of the Sacred Treasure by the Emperor of Japan in 1960, the National Medal of Technology in 1987, the Edison Award in 1989, and the Shewhart Medal of the American Society for Quality and Control in 1955. He was a fellow of the American Statistical Association and served as its president in 1945, a fellow of the Institute of Mathematical Statistics, and an honorary fellow of the Royal Statistical Society. The W. Edwards Deming Institute, chaired by his daughter Diana Deming Cahill, was incorporated in November 1993 to fulfill Deming's wish that his teachings be preserved intact and not be lost in the general flow of Total Quality Management (TQM). Columbia University has a Deming Center where students are taught his methods. Fordham University has the Deming Scholars MBA Program, an 18-month program that integrates Deming's principles with the conventional MBA curriculum. For those programs, for his ideas and teaching and the profound changes they have caused, and for the many areas where changes of this kind are still needed, will Deming be remembered for a long time. In 1922, Deming married Agnes Bell, who died in 1930. Their adopted daughter, Dorothy, died in 1984. In 1936, Deming married Lola Elizabeth Shupe, a mathematician, who died in 1986. He is survived by his daughters, Diana Deming Cahill and Linda Deming Ratcliff, five grandchildren, and three great-grandchildren. Michael J. Tortorella is a researcher at AT&T Bell Laboratories. 1994 American
Society for Quality |
