All Thinkers

W. Edwards Deming

William Edwards Deming (1900-1993) was an American statistician and management consultant whose work on quality control and systematic thinking about production reshaped manufacturing in Japan after the Second World War and, later, in the United States. He was born in Sioux City, Iowa, and grew up in a small town in Wyoming under difficult family circumstances. He studied electrical engineering at the University of Wyoming, earned a master's degree in mathematics and physics at the University of Colorado, and completed a PhD in mathematical physics at Yale in 1928. He worked for the United States Department of Agriculture and then the Census Bureau, where he applied statistical methods to sampling and the design of surveys. In 1947 he was invited to help prepare the Japanese census and returned to Japan in the early 1950s at the invitation of the Japanese Union of Scientists and Engineers. His lectures on statistical quality control and his broader management philosophy were widely adopted by Japanese industry, where he became a famous and revered figure decades before his ideas were taken seriously in his own country. The Deming Prize, established in Japan in 1951 and still awarded annually, recognised his influence. In the United States his work was largely ignored until a 1980 NBC documentary, If Japan Can, Why Can't We, brought him to public attention at the age of eighty. He spent his final thirteen years teaching, consulting, and writing; his major book Out of the Crisis appeared in 1982. He continued leading seminars until shortly before his death in 1993, aged ninety-three.

Origin
United States
Lifespan
1900-1993
Era
20th century
Subjects
Management Quality Control Statistics Manufacturing Systems Thinking
Skills
Scientific Thinking Critical Thinking Research Skills Ethical Thinking Problem Solving
Why They Matter

Deming matters because he showed that the quality of what an organisation produces depends primarily on the system of production, not on the effort or skill of individual workers, and that improving the system requires statistical understanding of the variation it produces. This view was radical. Most managers, then and now, respond to problems by exhorting or blaming workers, setting targets, or offering incentives. Deming's statistical analyses showed that most variation in output comes from the system — the equipment, procedures, materials, training, and management practices — not from the workers, and that attempts to improve quality by pressuring workers who have no control over the system produce worse results, not better. His quality approach, which emphasised continuous improvement, attention to variation, and the removal of systemic obstacles to good work, was adopted enthusiastically by Japanese industry in the 1950s and 1960s. By the 1970s Japanese manufacturers were producing goods of a quality that American industry could not match, in cars, electronics, and many other sectors. The American rediscovery of Deming in 1980 came at a moment of economic crisis when American firms were losing market share to Japanese competitors. His lessons — pay attention to variation, improve the system not the individuals, drive out fear, eliminate quotas and stack-ranking — transformed American manufacturing over the following decades and continue to influence management thinking. His legacy is visible in any organisation that takes process improvement seriously, from Toyota to hospitals using statistical techniques to reduce medical errors.

Key Ideas
1
Ninety-four percent of problems come from the system, not the worker
Deming's statistical studies showed that most variation in the quality of output comes from the system in which workers operate — the equipment, procedures, training, materials, management practices — rather than from the workers themselves. He sometimes gave the figure ninety-four percent system, six percent worker, based on his experience across many industries. This inverted the common managerial instinct of blaming workers when something goes wrong. If most problems come from the system, then telling workers to try harder will not fix them. Fixing them requires changing the system, which is the responsibility of management. The figure of ninety-four percent was a rough estimate, not a precise claim, but the underlying point remains sound and has been confirmed by subsequent work across many industries.
2
The Plan-Do-Study-Act cycle
Deming, adapting a method developed by his mentor Walter Shewhart, popularised a systematic approach to improvement: plan a change, do it on a small scale, study the results, and act on what you have learned (adopting, modifying, or abandoning the change). This cycle is repeated continuously as part of ongoing improvement. It sounds simple but is widely violated. Organisations often implement big changes without small-scale testing, declare victory without studying outcomes, or refuse to abandon approaches that are not working. The PDSA cycle, sometimes called the Deming Cycle, is the procedural backbone of quality improvement. It has been adopted across manufacturing, healthcare, education, and many other fields.
3
Drive out fear
Among Deming's famous fourteen points for management is the instruction to drive out fear. He argued that fear makes organisations worse at everything that matters. Fearful employees hide problems rather than reporting them, avoid taking initiative, do not share useful information with colleagues, and produce poor results while appearing compliant. Fear comes from blame, from stack-ranking systems that require some employees to be judged poor performers, from managers who punish bad news, and from uncertainty about job security. Deming did not propose soft management; he proposed intelligent management that recognised fear as a major source of dysfunction. Removing fear was a prerequisite for the honest communication that systemic improvement requires.
Key Quotations
"In God we trust; all others must bring data."
— Attributed, widely quoted
This saying captures Deming's insistence on evidence. Claims made without data — about quality, about productivity, about what works — should not be trusted just because someone makes them confidently. Management by opinion, by hunch, by authority, by tradition all produce the same problem: people defending positions with no grounds beyond assertion. Deming argued that serious management required serious measurement, and that conversations about what to do should proceed from data rather than from preferences. The saying has the playful tone of much of his teaching — half joke, half serious — and has spread far beyond its original context into any setting where evidence-based decisions matter.
"Ninety-four percent of problems are system problems, six percent are worker problems."
— Out of the Crisis, 1982
The specific percentage was rough, and Deming acknowledged as much. The point behind it was not. Most problems in most organisations come from the way the organisation is set up — the procedures, equipment, training, materials, and management practices — not from the workers who have to operate within it. Managers who respond to problems by blaming workers are addressing the small part rather than the large part. The quotation states the ratio so dramatically because the ordinary managerial instinct runs in the opposite direction. Changing this instinct was one of Deming's main goals. The ratio is worth remembering every time one is tempted to blame an individual for an organisational outcome.
Using This Thinker in the Classroom
Scientific Thinking When introducing the idea that systems matter more than individuals
How to introduce
Ask students: if a whole class is doing poorly on a topic, whose fault is that? Most will hesitate between the students and the teaching. Introduce Deming's view that most variation in output — whether factory defects or exam scores — comes from the system, not from individual effort. Discuss what this means for how we respond to problems. If the system is the main cause, then exhorting individuals to try harder will not fix things. Fixing things means changing the system — the teaching, the materials, the time allocated. Connect to the broader skill of distinguishing between system problems and individual problems, which applies far beyond business.
Critical Thinking When discussing the limits of incentive-based motivation
How to introduce
Tell students about Deming's red bead experiment: workers drawing beads from a box, where the proportion of red beads they draw is fixed by the contents of the box, not by their skill. Explain how the manager in the exercise praises, blames, and rates workers anyway, and how the workers' outputs scatter randomly regardless. Ask students: what is the experiment trying to show? Discuss the broader point. When output depends on the system rather than on worker effort, rewarding good performance and punishing bad performance is management theatre. It feels like action but does not change anything. Connect to contemporary debates about performance-related pay, grade-based incentives, and targets in schools.
Further Reading

For a short introduction: The New Economics for Industry, Government, Education (1993, MIT Press) is Deming's most accessible major book. The Out of the Crisis (1982, MIT Press) is more detailed but remains readable. The W. Edwards Deming Institute maintains an extensive online archive including video of his seminars and short introductory materials. Mary Walton's The Deming Management Method (1986) is a reliable and accessible biographical and conceptual introduction.

Key Ideas
1
Common cause and special cause variation
Deming distinguished between two kinds of variation in any process. Common cause variation is the ordinary fluctuation inherent in the system — the small differences from one unit of output to the next that arise from the countless small variables in any real production process. Special cause variation is a distinct event that takes the output outside the normal range — a machine breaking, a batch of bad materials, a worker error. The distinction matters because the two require different responses. Common cause variation is the signature of the system; reducing it means improving the system. Special cause variation is a specific event; responding to it means identifying and addressing the specific cause. Treating common cause variation as if it were special, or vice versa, leads to worse outcomes, not better.
2
Eliminate management by numerical goals
Deming argued strongly against management systems that set specific numerical targets without providing the means to reach them. If a plant is told to increase output by fifteen percent and given no new equipment, training, or resources, the only way to hit the target is to falsify the numbers, cut quality, or exhaust workers. All three happen. The targets become targets for appearing to perform, not targets for actually performing. Deming argued that the correct response to a problem is to study the system, identify the obstacles to good performance, and remove them — not to set a number and demand that workers meet it. This argument has been widely ignored in practice; much modern management still relies on numerical targets. The consequences Deming predicted have been widely observed.
3
The system of profound knowledge
In his later work Deming organised his thinking around what he called the system of profound knowledge, with four components: appreciation for a system, knowledge about variation, theory of knowledge, and psychology. A good manager needs all four. Appreciation for a system means understanding that the parts interact, that optimising one part often degrades another, and that the system must be managed as a whole. Knowledge about variation means understanding common and special cause variation. Theory of knowledge means understanding how we learn: through hypothesis, test, and revision. Psychology means understanding people — what motivates them, what discourages them, how they interact. The framework brings together statistical thinking, organisational theory, and human understanding into a single whole.
Key Quotations
"It is not enough to do your best; you must know what to do, and then do your best."
— Out of the Crisis, 1982
Deming is making a point that has wider application than business. Effort without understanding produces busy failure. The worker who tries hard to solve a problem they do not understand, the student who studies diligently but without grasping the material, the manager who pushes for results without knowing what would actually help — all are doing their best without knowing what to do. Deming argued that managers have a specific responsibility here: they cannot ask workers to do their best without first making sure that what is being asked is the right thing, and that the workers have the understanding needed to do it. The sentence is a gentle rebuke to the idea that effort alone is enough.
"A bad system will beat a good person every time."
— Attributed, widely quoted in Deming seminars
Deming is stating the practical implication of his systemic view of organisations. Put a capable, motivated, intelligent person into a dysfunctional system and the system will defeat them. They will work hard, produce poor results, become frustrated, and eventually either leave or give up. This is not because they lack capability; it is because the system is the larger force. The insight is worth holding against the common managerial reflex of trying to hire or train better people without addressing the system they will be working in. Changing individuals without changing the system produces the same outcomes, just with different individuals. The system change is the one that matters.
Using This Thinker in the Classroom
Scientific Thinking When examining variation as a signal about systems
How to introduce
Introduce Deming's distinction between common cause and special cause variation. Everyday examples help: your morning commute takes a different time each day (common cause variation arising from traffic, weather, your own preparation), but when it takes three times as long on a specific day (special cause — an accident). Ask students: should you respond to these differently? Discuss how responding to common cause variation as if it were special — trying to explain every small fluctuation by a specific cause — leads to bad conclusions. Discuss how missing special cause variation — writing off a real change as ordinary noise — leads to different bad conclusions. The skill is distinguishing the two.
Ethical Thinking When examining how fear affects organisational behaviour
How to introduce
Present Deming's instruction to drive out fear. Ask students: what happens in organisations where people are afraid? Discuss the ways fear distorts behaviour: people hide problems, avoid initiative, comply superficially while resisting substantively, and do not share useful information. Consider the sources of fear: blame, stack-ranking, job insecurity, unpredictable management behaviour. Ask: what would it take to reduce fear in a classroom, a workplace, a family? What is the difference between productive standards and fearful enforcement? Connect to broader discussions of how psychological safety affects what people are willing to say and do.
Critical Thinking When examining the danger of numerical targets
How to introduce
Present Deming's critique of management by numerical targets: setting a specific number that workers must hit without providing the means to hit it produces falsification, quality sacrifice, or exhaustion, but not genuine improvement. Ask students: can they think of examples from their own experience? Schools graded by test scores, hospitals measured by patient volume, police departments judged by arrest rates. In each case, what gets measured gets gamed. Discuss the deeper principle: a target without the system to achieve it is not a target, it is an instruction to produce the appearance of the target. Consider how genuine improvement requires working on the system, not just setting the number.
Further Reading

Andrea Gabor's The Man Who Discovered Quality (1990) is a full biography with substantive coverage of both his Japanese and American work. Henry Neave's The Deming Dimension (1990) provides a careful exposition of the fourteen points and the system of profound knowledge. For the Japanese context: the Japanese Union of Scientists and Engineers publishes materials on the Deming Prize and its history.

Key Ideas
1
The red bead experiment
Deming's famous red bead experiment was a classroom exercise in which volunteers were given a paddle with holes and asked to draw beads from a box containing white and red beads. The supposed task was to draw only white beads; any red bead was a defect. The proportion of red beads in the box was fixed — workers had no control over how many they drew. Deming, playing a tyrannical manager, would praise good performers, reprimand bad ones, set targets, and use stack-ranking. Over many rounds, the proportion of red beads per worker varied randomly around the fixed average. The experiment dramatised Deming's central point: when workers have no control over the system, managerial responses to their output are theatre, not management. The exercise was a staple of his seminars and remains an effective teaching device.
2
The funnel experiment
Another of Deming's famous teaching exercises, the funnel experiment, demonstrated what happens when you tamper with a stable process. A funnel is held above a target; a marble is dropped; where it lands is marked. The obvious improvement is to move the funnel to compensate for where the last marble landed. Deming showed that this standard reaction — adjusting the process after each result — makes the variation worse, not better. The marbles scatter more widely than if the funnel had been left alone. The exercise illustrated that responding to common cause variation as if it were special cause makes things worse. It also captures a common managerial dysfunction: the habit of constantly adjusting in response to random variation, mistaking noise for signal, and producing more noise as a result.
3
The seven deadly diseases
Deming identified seven deadly diseases of Western management, including lack of constancy of purpose, emphasis on short-term profits, annual performance ratings, excessive mobility of management, management using only visible figures, excessive medical costs, and excessive costs of warranty. The list captured structural weaknesses in the American corporate system of the 1980s. Some have improved since; others have grown worse. The critique of annual performance ratings was particularly sharp — Deming argued these systems were statistically meaningless because the differences between workers were mostly smaller than the noise from common cause variation, so the ratings measured randomness dressed up as judgment. Several large companies have since moved away from traditional rating systems, in part because of this argument.
Key Quotations
"The aim of leadership is not merely to find failure but to remove the causes of failure."
— Out of the Crisis, 1982
Deming is distinguishing between two conceptions of management. One defines management as detecting and punishing failures when they occur. The other defines it as removing the conditions that produce failures in the first place. The first is reactive and looks for someone to blame. The second is systemic and looks for what can be changed. Both require attention to where failures occur, but they respond differently. Deming's preferred conception treats failures as symptoms of systemic problems that management is responsible for addressing. The worker who made the error is usually less important than the process that allowed the error. The distinction shapes what kind of leader a manager becomes.
"Without data, you're just another person with an opinion."
— Attributed to Deming, widely quoted
This sharp version of Deming's methodological point is often attributed to him, though its exact wording may be a student's paraphrase. The claim is pointed. In most organisational settings, the person with the strongest views often dominates conversations about what to do. Authority and assertiveness count for more than evidence. Deming argued against this norm: without data, no one's opinion is especially privileged. The critique cuts against hierarchical management culture in which senior people's preferences are treated as decisive. The alternative Deming proposed required bringing data into conversations and treating all participants as subject to correction by evidence. This is harder than it sounds; most organisations resist it.
Using This Thinker in the Classroom
Ethical Thinking When examining annual performance ratings
How to introduce
Introduce Deming's argument against annual performance ratings: that the differences between workers are mostly smaller than the statistical noise in their outputs, so ratings measure random variation as much as real difference. Ask students: what are the implications? Discuss how rating systems affect trust, cooperation, and willingness to take risks. Consider recent moves by companies away from stack-ranking toward other forms of performance management. What does it mean to judge individual performance fairly when most variation comes from the system? How should schools think about the same problem — about how much exam variation reflects teacher, school, or student effort versus systemic factors?
Cultural Heritage and Identity When examining why ideas spread in one country but not another
How to introduce
Tell students that Deming's ideas were adopted in Japan in the 1950s, where they transformed manufacturing, but were largely ignored in the United States until the 1980s, when American industry was in crisis. Ask: why this pattern? Discuss possible reasons — Japanese openness to outside ideas after the war, American confidence in its own methods, the specific role of the Deming Prize in building institutional commitment in Japan. Consider the broader question: how do new ideas spread between countries and industries? What makes one society receptive to a given idea and another resistant? Connect to the history of how scientific and managerial knowledge has moved around the world.
Common Misconceptions
Common misconception

Deming's ideas are mainly about statistics and apply only to manufacturing.

→
What to teach instead

The statistical methods were only part of Deming's work, and he insisted that understanding statistics was not the main thing. The main thing was understanding organisations as systems — the interplay of management practices, worker psychology, variation, and learning. Deming's later work explicitly applied the same framework to service industries, healthcare, education, and government. The reduction of his thinking to statistical process control happened partly because that was easier to teach and market than his broader organisational philosophy, and partly because the implications of the broader philosophy (drive out fear, eliminate numerical targets, abolish annual ratings) are more difficult to implement than running control charts. The narrower version is a real part of his work but is not the whole.

Common misconception

Deming opposed measurement in management.

→
What to teach instead

Deming was one of the strongest advocates of measurement in the history of management thought. His criticism was not of measurement but of the misuse of measurement — particularly of setting numerical targets without providing the means to achieve them, and of using ratings that did not distinguish signal from noise. His positive position was that good measurement requires statistical understanding, careful distinction between common and special cause variation, and attention to what measurements actually track. Confusing his critique of misused measurement with opposition to measurement itself misses his actual position. He wanted more and better measurement, used more intelligently — not less measurement.

Common misconception

Deming's Japanese success shows that his methods work only in collectivist cultures.

→
What to teach instead

The cultural explanation of Deming's Japanese success has some merit but is often overstated. His methods were also adopted successfully by American firms after 1980 — Ford, Xerox, and many others implemented quality approaches drawn directly from Deming and achieved significant improvements. The methods have since been adopted in service industries across many different cultures. What the Japanese case shows is that a society that took Deming's ideas seriously in the 1950s was able to build competitive advantage, while the United States paid little attention for three decades and lost ground accordingly. The lesson is about paying attention, not about culture.

Common misconception

Total Quality Management and Six Sigma are straightforward applications of Deming's ideas.

→
What to teach instead

TQM and Six Sigma, as they were commonly implemented, often departed significantly from Deming's actual views. Many TQM programmes in the 1980s and 1990s focused on tools and techniques while ignoring the systemic and psychological aspects Deming emphasised — keeping numerical targets, annual ratings, and fear-based management while adding statistical tools on top. Six Sigma, which prescribes the elimination of variation to specific defect rates, is compatible with aspects of Deming's thinking but is often applied as a target-driven programme of the kind he warned against. Deming himself was critical of several prominent programmes launched in his name. Associating his work too quickly with any specific subsequent programme risks missing what he actually argued.

Intellectual Connections
Complements
Peter Drucker
Deming and Drucker were near-contemporaries who approached management from different directions but reached convergent conclusions. Drucker came from economics and philosophy, analysing what organisations are for and how they should be structured. Deming came from statistics, analysing how production processes actually produce variation and quality. Both rejected authoritarian management, both insisted on treating workers as capable adults, and both thought that good management was about understanding systems rather than about controlling individuals. Reading them together shows how the serious study of management developed across the twentieth century, with different methods converging on overlapping practical conclusions.
Complements
Taiichi Ohno
Ohno at Toyota built a production system — later called lean manufacturing or the Toyota Production System — that shared many features with Deming's recommendations: attention to variation, worker participation in process improvement, elimination of waste, continuous refinement. Deming's ideas had reached Japan before Ohno's system reached maturity, and the two thinkers were part of an overlapping Japanese engineering and management community. Reading them together shows how a coherent approach to quality and production developed in Japanese industry, with contributions from American, Japanese, and other sources. The result transformed global manufacturing.
In Dialogue With
Elinor Ostrom
Ostrom's work on how communities self-organise to manage shared resources shares with Deming the conviction that good institutional outcomes depend on the design of the system of relations, not just on the effort or character of participants. Ostrom studied fisheries, irrigation systems, and other commons; Deming studied production systems in factories. Both showed, in their different domains, that well-designed systems outperform badly-designed ones regardless of the moral or intellectual qualities of the individuals in them. Reading them together shows how institutional design has emerged as a serious subject across several disciplines.
In Dialogue With
Paulo Freire
Freire's pedagogy of the oppressed argued against top-down teaching methods in which students are treated as empty vessels to be filled by authoritative teachers. Deming's management philosophy argued against top-down management methods in which workers are treated as passive executors of management decisions. The parallel is not accidental. Both thinkers saw that real improvement — in learning or in production — requires the active participation of the people whose work is being directed, and that treating them as passive objects produces poor outcomes. Reading them together reveals a broader twentieth-century critique of top-down expertise that reshaped multiple fields.
Anticipates
Paul Farmer
Farmer's work in global health argued that the failures of healthcare in poor countries were systemic failures — not the fault of individual doctors or patients but of the underlying systems of resource allocation, training, and supply. Deming's earlier work in industry had argued the same thing about manufacturing failures. Farmer in healthcare applied an analogous systems-level analysis, with analogous practical implications: to fix a system, work on the system, not on blame for individuals. Reading them together shows how the basic insight about systemic responsibility has been applied across very different domains to produce real improvements in lives and outcomes.
Complements
Mary Parker Follett
Follett's collaborative management philosophy, developed in the 1920s, anticipated many of Deming's mid-century arguments. Both rejected command-and-control management; both emphasised the importance of working with employees rather than on them; both saw conflict as something to be used productively rather than suppressed. Deming was aware of Follett's work and drew on its broad orientation. Reading them together reveals a continuous line of thought about management-as-collaboration that was minority opinion throughout much of the twentieth century and that has become more influential in recent decades.
Further Reading

For scholarly depth

The journal Quality Progress has published extensive work on Deming and his reception. John Dowd's The Deming Route to Quality and Productivity (1986) is a technical treatment. The archives at the US Library of Congress hold substantial Deming materials including correspondence, lectures, and unpublished writings.

Rafael Aguayo's Dr

Deming

The American Who Taught the Japanese About Quality (1990) is a detailed and reliable account of the Japanese period.