W. Edwards Deming, an expert on business management who advised Japan on how to rebuild its shattered industries after World War II and urged American corporations to treat their workers as associates rather than adversaries, died early yesterday at his home in Washington. He was 93.
The cause was cancer. Although he was ill in recent years, Mr. Deming continued to work, conducting the last of his four-day seminars on quality-management in the Los Angeles area from Dec. 7 through Dec. 10.
Mr. Deming's theories were based on the premise that most product defects resulted from management shortcomings rather than careless workers, and that inspection after the fact was inferior to designing processes that would produce better quality.
He argued that enlisting the efforts of willing workers to do things properly the first time and giving them the right tools were the real secrets of improving quality -- not teams of inspectors. Initial Success in Japan
Mr. Deming was an obscure statistician in this country in 1950 when research he had conducted during World War II came to the attention of some Japanese industrial leaders. At their request, he then gave a series of lectures in Japan on his quality-control principles, and he and his message were eagerly embraced. The Japanese, who lacked many natural resources or a colonial empire, were a receptive audience because they believed they would prosper only if they could sell products on world markets.
His advice to Japan made Mr. Deming the leader of a generation of specialists on product durability and reliability who were then sought by American companies trying to catch up to Asian competitors. But his renown in the United States never matched the reputation he achieved in Japan.
After the application of his methods brought enormous commercial success to some Japanese companies, the Japanese created a Deming Prize for companies that made striking advances in quality. Mr. Deming was described by many commentators as the best-known and respected American in postwar Japan after Gen. Douglas MacArthur.
William Edwards Deming was born on Oct. 14, 1900, in Sioux City, Iowa, to a family whose roots in America reached back to the Revolutionary War. He grew up in strained financial circumstances in Powell, Wyo., where his father worked as a part-time lawyer and land developer.
Mr. Deming attended the University of Wyoming in Laramie, working at odd jobs while he studied engineering. He later earned a master's degree in mathematics and physics from the University of Colorado and was awarded a doctorate in physics from Yale University in 1928. Management as Problem
Mr. Deming used the later years of his long career to try to reform American management, for considerable fees, sometimes as much as $100,000 a year from a single client. A tall, formal man who habitually wore frayed three-piece suits and spoke to senior executives as if they were schoolboys, he delighted in telling corporate chieftains who asked him to help solve a company's problems that they were a significant part of the problem.
"Can you blame your competitor for your woes?" he would intone to groups of corporate managers. "No. Can you blame the Japanese? No. You did it yourself."
Although the core of his method to improve quality was the use of statistics to detect flaws in production processes, he developed a broader management philosophy that emphasized problem-solving based on cooperation. He exhorted managers to "drive out fear," so that workers would feel free to make improvements in the workplace.
Mr. Deming denounced management procedures like production quotas, performance ratings and individual bonuses, saying they were inherently unfair and detrimental to quality. He said customers would get better products and services when workers were encouraged to use their minds as well as their hands on the job.
A frugal man in his personal life, Mr. Deming drilled companies to work relentlessly to reduce waste -- anything from parts sitting unused in inventory to motions by a worker that did not add value to the final product. One of his daughters recalled that he had dated the eggs in his refrigerator with a felt-tipped pen so that the oldest would be eaten first and none would go to waste. Success at Ford and Xerox
His blunt approach offended many executives, who turned elsewhere for advice. Yet some companies, including the Ford Motor Company and the Xerox Corporation, sent hundreds of their top-level managers to his lectures and seminars. "He said the only way to bring about change was to have direct contact with senior management," said James K. Bakken, a former vice president at Ford.
Well into his 90's, Mr. Deming maintained an active travel schedule, crisscrossing the country to conduct seminars and consult with companies he considered sufficiently motivated to benefit from his attention. He also lectured at Columbia University's Business School and taught continuously at New York University's Stern School of Business from 1946 until the end of the spring term this year.
Integral to Mr. Deming's approach was an emphasis on sensitivity to customer needs. Everyone has a customer, he said, either inside or outside the organization. Although he developed his methods in an industrial setting, Mr. Deming insisted that his approach was applicable to institutions generally, even those in service and nonprofit businesses.
After Mr. Deming's first trip to Japan, the Japanese Union of Scientists and Engineers assembled his lecture notes and published them as a book, "Elementary Principles of the Statistical Control of Quality." Prize Established by Japanese
Mr. Deming refused to accept royalties from the book and suggested that the funds be used to promote quality. The group thus established the Deming Prize for achievements in quality, which quickly became one of the awards most sought by Japanese companies.
Yet Mr. Deming remained little known in the United States, where wartime efforts to establish statistical quality-control methods fell before the headlong rush to push products out of factories. The prevailing sentiment in American industry from the 1950's until the late 1970's was that more quality meant more cost and that consumers did not want to pay for higher-quality products.
That attitude began to change when Japanese products with brand names like Sony and Panasonic drove the American consumer-electronics industry almost out of business while reliable, fuel-efficient Toyotas and Hondas gnawed away at the domestic auto industry.
One of the first large American corporations to seek Mr. Deming's help was Ford Motor. Ford officials persuaded him to visit their headquarters in Dearborn, Mich., in February 1981, when the company's sales were faltering and it was losing hundreds of millions of dollars.
Ford executives were expecting a slick presentation on tricks to improve quality. Mr. Deming, instead, insisted on questioning the company's culture and management philosophy. Eighty-five percent of quality problems, he told them, are the result of management errors.
"We were sitting there with our pens poised to write down the prescription for what we should do about quality," Mr. Bakken recalled. "The first thing he said was, 'Do you have a constancy of purpose?' We were not quite sure what to make of him."
Because Mr. Deming was sponsored by Donald E. Petersen, Ford's president at the time, the relationship survived the executive egos bruised by these early encounters, and statistical control charts blossomed in the company's factories. In the 1980's, Ford led the domestic auto industry in quality improvements. No Formal Organization
As Ford's success became obvious, demand for Mr. Deming's services grew. He kept his client list short and refused to have anything to do with companies not willing make top executives available to him. Among the companies that turned to Mr. Deming and his disciples were Dow Chemical, Procter & Gamble, American Telephone & Telegraph and The New York Times.
Unlike other quality experts, like Joseph Juran and Philip Crosby, Mr. Deming never built a formal organization. He continued to work as a solo practitioner out of an office in the basement of his modest home in Washington. But he did develop an informal alliance of followers and often required clients to hire a member of his circle to teach statistical methods and instill his philosophy.
Mr. Deming developed that philosophy in the 1920's and 30's while working at A.T.& T.'s Hawthorne manufacturing plant in Chicago and as a protege of Walter Shewhart of Bell Laboratories. Mr. Shewhart was a pioneer in the use of statistics to control manufacturing processes.
Although his academic training was in mathematics and physics, Mr. Deming had mastered statistical theory and practice, in part by taking a one-year leave of absence from the Agriculture Department in the mid-1930's to study under a pioneer of the discipline, Sir Ronald Fisher of the University of London.
In the 1930's, Mr. Deming helped design the sampling techniques used by the Census Bureau. And in World War II he helped military planners apply statistics to the production of supplies for warfare. Changing the Culture
Companies that sought to improve their quality by adopting Mr. Deming's methods often found they had to change their entire culture. To convince workers that managers really did want to enlist them as partners, many companies eliminated cherished management perquisites like special parking spaces and executive dining rooms because shop-floor workers found them offensive.
One Deming lesson to designers and engineers was to change the way they thought about quality. Traditionally, specifications for almost any physical object were set at a desired value with a certain allowable deviation. Anything within the deviation limits was considered a good part, anything outside the limits bad.
But Mr. Deming's statistical studies showed that with complex products, these allowable deviations could add up to a defective final product. The message to managers was that simply being within specification was not good enough. Processes must be continually honed, he taught, to reduce the variability from part to part.
Mr. Deming's first wife, Agnes, died many years ago. His second wife, Lola, died in 1986. He is survived by two daughters, Diana D. Cahill of Palos Verdes, Calif., and Linda D. Ratcliff of Potomac, Md.; five grandchildren, and three great-grandchildren.
