Out of the Crisis

W. Edwards Deming, a statistician and management consultant whose work helped transform Japanese industry after World War II, argues that American industry faces a crisis rooted not in foreign competition or insufficient technology but in the failure of management itself. The book lays out a comprehensive theory of management transformation built around Deming's 14 Points for Management, fourteen principles for organizational transformation, and a diagnosis of what he calls the "deadly diseases" afflicting Western companies. Deming contends that neither harder work, new machinery, nor slogans can halt the decline of American industry. Only a fundamental change in the philosophy and practice of management, sustained over decades, can restore competitive position. The book makes no distinction between manufacturing and service industries, asserting that both are subject to the same principles.

Deming opens by challenging what he considers a pervasive American misconception: that quality and productivity are incompatible. Drawing on his experience teaching Japanese executives beginning in 1950, he presents a chain reaction in which improving quality leads to decreased costs, improved productivity, greater market share, and more jobs. Japanese management, guided by the statistical methods of Walter A. Shewhart, a Bell Telephone Laboratories physicist and statistician, and by Deming's own teaching, adopted this chain reaction as a guiding principle. To illustrate, Deming presents a factory where management developed operational definitions, meaning explicit testable criteria for acceptable and unacceptable work, and the defective rate dropped from 11 percent to 5 percent at zero additional cost. A second example from the Nashua Corporation shows how achieving statistical control of a paper-coating process, meaning a stable and predictable state of measured variation, saved $800,000 per year and opened the way to further engineering innovation.

The core of the book is the 14 Points. Point 1 calls for constancy of purpose, urging companies to plan for long-term survival rather than chasing quarterly dividends. Point 2 demands a new philosophy that refuses to tolerate commonly accepted levels of mistakes and defects. Point 3 argues that mass inspection is too late and costly, since quality must be built into the product. Point 4 attacks the practice of awarding business to the lowest bidder, advocating instead long-term relationships of loyalty and trust with single suppliers. Point 5 calls for constant improvement of the system, with quality built in at the design stage. Point 6 addresses the need to reconstruct training, and Point 7 redefines supervision as leadership aimed at helping people and machines do better work.

Point 8 calls for driving out fear, which Deming identifies as one of the most corrosive forces in organizations: fear of job loss, of putting forth ideas, of admitting mistakes, and of the annual performance rating. Point 9 urges breaking down barriers between departments. Point 10 calls for eliminating slogans and exhortations directed at the workforce, arguing that most causes of low quality belong to the system and lie beyond workers' power to fix. Points 11a and 11b call for eliminating numerical quotas on the factory floor and numerical goals for management, respectively. Point 12 addresses barriers that rob people of pride in their work. Point 13 encourages education and self-improvement, and Point 14 calls for putting everyone to work on the transformation using the Shewhart cycle, a four-step iterative process of planning a change, carrying it out, observing the effects, and studying the results.

Deming then identifies the deadly diseases of Western management. The first is lack of constancy of purpose. The second is emphasis on short-term profits, driven by fear of unfriendly takeover and pressure from stockholders. Deming cites Yoshi Tsurumi, a scholar of American and Japanese management practices, who observed that most American executives believe they are in business to make money rather than to produce products and service. The third disease, and the one Deming attacks most forcefully, is the annual performance review. He calls it devastating: It nourishes short-term thinking, demolishes teamwork, and leaves people demoralized. Using statistical examples, he demonstrates that apparent differences between workers within a system are often attributable to the system itself rather than to individual ability. The fourth disease is mobility of management, which annihilates teamwork and long-term commitment. The fifth is running a company on visible figures alone, ignoring figures that are unknown and unknowable, such as the multiplying effect of a satisfied or dissatisfied customer. Among the obstacles Deming catalogs are hope for instant results, the belief that automation will solve everything, and the assumption that troubles lie entirely in the workforce. He cites quality expert Joseph M. Juran's observation that most possibilities for improvement lie in action on the system, not in worker effort.

The book's central technical contribution is the distinction between common causes and special causes of variation. Common causes belong to the system and can be removed only by management. Special causes are attributable to specific workers, machines, or events. Deming estimates that 94 percent of troubles belong to the system. He describes two types of mistake: treating a common cause as special, which leads to overadjustment, and treating a special cause as common, which leads to inaction. Shewhart's 3-sigma control limits, statistical boundaries that separate routine variation from signals requiring individual attention, provide a rational guide to minimize the cost of both mistakes. Simulated experiments with a funnel demonstrate that adjusting a stable process based on the last error doubles the variance, while more aggressive adjustment causes the system to explode. The red beads experiment, the chapter's centerpiece, uses six workers drawing beads from a bowl containing 20 percent red beads. Despite identical procedures, their results vary widely, yet all fall within the system's limits. Ranking them or rewarding the best performer would amount to a lottery.

Deming applies these principles across domains. A chapter on quality argues, drawing on Shewhart, that quality is determined not by the product alone but by the interaction among the product, the user, and the instructions for use. A chapter on service organizations extends the 14 Points to hospitals, banks, utilities, and government agencies, with case studies including a bank that used statistical methods to reduce errors and an electric utility whose quality circle, a small worker group organized to solve quality problems, predicted underground cable failures and saved hundreds of thousands of dollars. A chapter on operational definitions argues that terms like "safe" and "clean" have no communicable meaning until expressed as a specific test, a criterion, and a yes-or-no decision. A chapter on inspection theory proves that minimum total cost is achieved by either no inspection or 100 percent inspection, and that any intermediate sampling plan, a sample-based protocol for accepting or rejecting batches, can only increase total cost. Downstream examples reinforce that improvement is management's responsibility: In one case, statistical analysis of 150 truck drivers' mistakes revealed that 82 percent arose from the system, making the company's blanket warning letters counterproductive.

On the question of how long transformation will take, Deming is blunt: decades. He argues that one cannot catch up with someone who is continually gaining speed, and that the question may not be when but whether American industry will transform. He identifies retarding forces including management paralysis, government regulations, and antitrust enforcement that makes cooperative efforts between companies difficult.

The book closes with an appendix tracing the transformation in Japan. Deming recounts his resolution not to repeat the mistakes made in America during 1942 to 1948, when statistical methods flared and then died because management did not understand their responsibilities. The critical step was reaching Japanese top management through Ichiro Ishikawa, president of the Kei-dan-ren (the Federated Economic Societies) and of JUSE (the Union of Japanese Scientists and Engineers). Education expanded rapidly, and early results were dramatic: one company reduced rework to 10 percent of its former level, another tripled pharmaceutical output with the same resources, and a steel company reduced fuel consumption by 29 percent per ton. QC-Circles, small worker groups for quality improvement formalized by quality control expert Dr. Kaoru Ishikawa by 1960, represented the Japanese way of working together to eliminate special causes and improve the system. The transformation, Deming argues, happened because Japanese management accepted their responsibilities and never stopped improving.