Why We Need a New Theory of Organization Now

Theory X Was Right

Frederick Taylor published Principles of Scientific Management in 1911 and changed how organizations thought about work. His insight was structural: don’t make people work harder — redesign the work itself. Study the task, find the best method, train workers to follow it.

This gets caricatured as contempt for workers. It wasn’t. It was a rational response to real conditions. Most industrial workers in 1911 had limited formal education. The knowledge to optimize work genuinely resided with engineers and managers, not on the factory floor. The coordination problem at scale was enormous — hundreds of people performing physical, repetitive tasks in environments where deviation meant injury or waste. And motivation could reasonably be reduced to wages and job security because for most workers, that’s what was at stake.

Theory X — the assumption that workers need to be directed, controlled, and motivated through external rewards and consequences — wasn’t a character judgment. It was an operating model matched to its scarcity structure: knowledge was scarce and concentrated at the top, labor was abundant and interchangeable at the bottom. The management system that emerged — hierarchical control, prescribed methods, narrow task definitions, close supervision — was the architecture that scarcity demanded.

It worked. Not perfectly, not humanely by today’s standards, but it worked. It built the industrial economy.

Then the World Changed

World War II compressed a century of social evolution into five years.

The war was a mass technical education program. Millions of people who had never seen complex machinery learned to operate radar, maintain aircraft engines, decode encrypted communications, calibrate artillery, and run logistics across continents. They came home not just willing to handle complexity but proven in it. The knowledge asymmetry between management and labor — the structural foundation of Theory X — had narrowed dramatically.

The GI Bill finished what the war started. Nearly eight million veterans used it to attend college or vocational school. Before 1944, higher education was largely for the wealthy. By the early 1950s, factory supervisors were managing people with more formal education than they had. The assumption that management knows and workers execute no longer described reality.

Women entered the industrial workforce at scale. Six million women took skilled positions during the war, proving that the entire category of “who counts as a capable worker” had been structurally wrong. The pre-war management model assumed a homogeneous male workforce whose motivation reduced to wages. A diverse workforce with different circumstances and different relationships to work couldn’t be managed as interchangeable units, even in theory.

The economy shifted from production to innovation. Wartime manufacturing proved American industry could produce at extraordinary volume. The postwar challenge was different — product development, marketing, white-collar coordination, technical problem-solving. Work that required judgment, discretion, and initiative. Work that Theory X couldn’t govern.

The workforce was more educated, more technically skilled, more diverse, and less willing to be treated as interchangeable capacity. Theory X hadn’t become wrong in the abstract. The conditions that made it right had changed.

McGregor Saw It

Douglas McGregor didn’t invent Theory Y from idealism. He observed that management systems were producing the behavior they then pointed to as justification. Design for distrust and you get people who can’t be trusted — not because the assumption was right, but because the structure trained judgment out of them.

His argument was architectural: management systems create the behavior they assume. Theory X systems — approval chains, surveillance, narrow task definitions — produced avoidance and passivity. Not because workers were inherently lazy, but because the structure made passivity rational.

Theory Y was the better assumption for the postwar world: people are capable of self-direction, creativity, and commitment to objectives when the conditions support it. The key word is conditions. McGregor had tried pure aspiration at Antioch College, where he served as president from 1948 to 1954, and it hadn’t worked. “I thought I could avoid being a boss,” he later reflected. He learned the hard way that believing in people without redesigning the structure produces good intentions consumed by institutional reality.

So Theory Y, published in 1960, wasn’t a call to simply trust people more. It was a call to redesign the conditions of work — the measurement systems, the authority structures, the career incentives — so that human capability could actually express itself. The structure is the riverbed; behavior is the current. Move the rock, and the water follows.

Theory Y was right for its era. The humanistic management tradition it spawned — participative leadership, management by objectives, self-directed teams, engagement-driven cultures — matched a world where the binding constraint was unlocking human potential within organizations that had been designed to suppress it.

Now the World Is Changing Again

AI changes what’s scarce.

For sixty years, Theory Y has operated within the same fundamental scarcity structure McGregor inherited: human labor and expertise as the binding constraints. Whether you managed people through control (Theory X) or through empowerment (Theory Y), the central problem was the same — how to organize, develop, and deploy scarce human capability. Theory Y said do it by enabling people. Theory X said do it by directing them. Both agreed on what was scarce.

AI inverts the scarcity. Execution, retrieval, drafting, classification, routine analysis, first-line decision support — the cognitive work organizations have spent decades hoarding, structuring, and rationing — become abundant. What becomes scarce is judgment, context, trust, governance, and the ability to design systems that use abundant intelligence well.

This is not a technology shift. It is a scarcity shift, and scarcity shifts change which management theory is right.

A Theory Y organization that adds AI to its existing operating model — copilots within existing workflows, chatbots on existing service models, automation of existing processes — will produce local efficiency gains and zero structural advantage. The operating model is still organized around scarce human capacity. The AI is just making that model run faster. A better interface to the same architecture.

Theory A

We need a new theory of organization because AI changes the structural conditions as profoundly as World War II did. Theory A — Agency, Autonomy, Architecture, Agentic — is the management theory for the post-AI scarcity structure.

The argument is McGregor’s, extended: structure determines behavior. But the structure must now be designed for a world where intelligence is abundant and judgment is scarce. That means:

Outcomes replace activities. When AI can execute, the operating model must define what conditions must be true — not what work must be performed. “Employees are paid accurately and on time” rather than “process payroll within cycle.” Accountability for outcomes frees the path; accountability for activities prescribes it.

Signals replace surveys. The organization already emits continuous evidence of its own condition — case data, exception rates, cycle times, behavioral patterns. Reading this as intelligence rather than workload is the single largest source of new value. Periodic surveys measuring stated preference cannot match what emitted signals reveal about reality.

Constraints replace processes. Standardize the boundaries — what must always be true, what cannot be violated, which decisions require human judgment — and let the path personalize. This is how you govern AI autonomy at scale: not by prescribing every step, but by defining the riverbed within which the current flows safely.

Management unbundles. The traditional manager bundled operational direction with career stewardship in one role because there was no other way. AI makes that bundling an obstacle: people flowing to where judgment is needed can’t change managers every time, and operational urgency always crowds out human development when one person owns both. Separate who directs the work from who develops the people.

Self-actualization becomes the baseline. Maslow’s hierarchy placed self-actualization at the peak — a luxury available after the lower needs were met. In an AI-native organization, the routine work that occupied the lower levels is absorbed by machines. What remains for humans is judgment, creativity, stewardship, and design. Self-actualization isn’t the peak you climb to. It’s the operating condition the structure must support from day one.

The Pattern

Theory X was right for a world of scarce knowledge and abundant labor. The war changed the workforce — more educated, more skilled, more diverse, less willing to comply — and Theory X became an obstacle.

Theory Y was right for a world of scarce human capability and abundant ambition. AI changes the scarcity — intelligence becomes abundant, judgment becomes scarce — and Theory Y, while not wrong, becomes incomplete. It still assumes human capacity is what must be unlocked. The new question is how to design organizations where human judgment and machine intelligence amplify each other.

Management theories don’t fail because their authors lacked insight. They fail because the conditions that made them right evolve. The conditions are evolving now, faster than at any point since 1945. The organizations that recognize the scarcity inversion and redesign their operating models accordingly will build structural advantage that compounds. Those that add AI to existing models will get efficiency gains that plateau.

The riverbed is the operating model. The current is the behavior. The rock has shifted. It’s time to redesign.