1

Something has gone wrong inside the modern corporation, and it happened so gradually that almost nobody noticed. Over the past three decades, the dominant response to every operational problem—safety incidents, quality failures, regulatory risk, earnings misses—has been the same: add a layer of technology, add a layer of compliance, remove a layer of human judgment. Automate the decision. Monitor the employee. Replace discretion with a dashboard.

The logic feels irresistible. Technology is consistent. Technology doesn’t get tired. Technology doesn’t cut corners on a Friday afternoon. And so, one system at a time, the modern workplace has been engineered to minimize the role of the individual human being. The worker’s job is increasingly to serve the system rather than the other way around. Decisions that were once made by a foreman on a factory floor are now made by an algorithm in a server farm. The foreman is still there. He just doesn’t matter as much.

This is the escalating removal of personal agency—the individual’s sense that what they do, decide, and observe actually matters to the organization. And it carries a cost that never appears on a balance sheet but shows up everywhere else: in disengagement, in turnover, in the slow erosion of the discretionary effort that separates a functioning company from an excellent one. When people do not believe their judgment is valued, they stop offering it. When they stop offering it, the organization loses the single most valuable source of operational intelligence it possesses: the observations of the people closest to the work.

In 1987, a quiet government bureaucrat named Paul H. O’Neill walked into exactly this kind of organization—a hundred-year-old Rust Belt industrial giant called Alcoa—and did the opposite of what every modern management instinct would prescribe. Instead of layering on compliance and technology, he restored personal agency. And in doing so, he triggered a cascade that ran from employee satisfaction through customer satisfaction, from efficiency through productivity, from EBITDA improvement through enterprise value creation, and ultimately to one of the most remarkable corporate transformations in American history.

2

When O’Neill became chief executive of Alcoa, the Aluminum Company of America, the company was hemorrhaging market share to subsidized foreign producers. Its diversification strategy had failed. Its leadership had been ousted in a quiet boardroom coup orchestrated in part by fellow director Alan Greenspan.¹ Alcoa was a textbook candidate for the compliance playbook: hire a restructuring firm, cut headcount, install monitoring systems, and report progress in quarterly decks full of traffic-light charts.

O’Neill did none of this. He stood before a ballroom of Wall Street analysts and told them he intended to make Alcoa the safest company in America. His goal was zero injuries.²

One investor reportedly bolted for a payphone and told his clients the board had put a “crazy hippie” in charge.³ That investor’s clients missed a 700 percent increase in stock price, an 818 percent increase in market capitalization, and a sevenfold increase in net income over the next thirteen years.⁴

The investor made a rational mistake. He heard safety and thought compliance—the kind of initiative that costs money, produces binders, and changes nothing fundamental. What O’Neill was actually proposing was the restoration of personal agency to every one of Alcoa’s 59,000 employees, using safety as the mechanism and incentive alignment as the architecture.

3

To understand why O’Neill’s approach worked, you have to understand the two models available to any leader facing a distressed organization.

The compliance model says: people are the problem. Write rules to constrain them. Deploy technology to monitor them. Punish violations. This model reduces the worst outcomes. It also systematically removes personal agency, because it tells every employee, in effect: we do not trust your judgment. Follow the procedure. If the procedure is wrong, that is not your problem.

The result is an organization full of people who do exactly what is required and nothing more. They do not volunteer observations. They do not suggest improvements. They do not take ownership of outcomes, because the system has explicitly told them that outcomes are owned by the process, not by them. Compliance produces a floor. It cannot produce a ceiling.

The incentive-alignment model says: people are the solution. Find the one goal that serves everyone’s interest—labor, management, and capital. Build the infrastructure for them to pursue it. Remove the barriers that prevent them from acting on what they see. The result is the opposite of compliance: it is an organization full of people who believe their judgment matters, because the system has been explicitly designed to reward the exercise of that judgment.

This is the restoration of personal agency. And personal agency, once restored, does not confine itself to a single topic. A worker who is empowered to stop an unsafe machine today is a worker who suggests a more efficient workflow tomorrow, identifies a quality defect next week, and improves a customer’s experience next month. The habit of voice, once established, becomes the operating system of the entire enterprise.

4

O’Neill built three interlocking mechanisms, each designed to restore personal agency and align incentives simultaneously.

Mechanism One: Transparency Replaces Surveillance

O’Neill required every facility to report every safety incident to headquarters within twenty-four hours, with a root-cause analysis and corrective action.⁵ This sounds like compliance. It was not. The distinction is in the direction of the incentive. In a compliance system, the worker reports because they fear punishment for not reporting. In O’Neill’s system, the worker reported because reporting was rewarded—managers who surfaced problems received resources to fix them. Managers who concealed problems were fired.

When a senior executive at an Alcoa plant in Mexico covered up a carbon monoxide incident that injured 150 workers, O’Neill terminated him within two days. The reason: not the incident itself, but the concealment. “No one else had the opportunity to learn from it.”⁶

This inverted the information architecture of the entire company. In a compliance culture, the rational individual conceals. In an agency culture, the rational individual discloses—because disclosure is how resources flow toward problems. Technology was used not to monitor employees but to amplify their voice: O’Neill commissioned Carnegie Mellon graduates to build a real-time information system that carried safety data from the factory floor to the CEO’s desk in hours, not weeks.⁷ The technology served the worker’s agency. The worker did not serve the technology.

Mechanism Two: Authority Replaces Permission

O’Neill gave his personal phone number to hourly factory workers and told them to call him directly if their supervisor ignored a safety concern.⁸ He told management: “As soon as anyone identifies a condition that could hurt someone, fix it. We are not going to budget for safety. I will figure out how to pay for it.”⁹

Consider what this did to personal agency. In the old Alcoa, a factory worker who saw a dangerous condition had to fill out a form, submit it to a supervisor, wait for budget approval, and hope the request survived three layers of management review. The worker’s agency was zero. In O’Neill’s Alcoa, that same worker could stop the line, call the CEO, and watch the problem get fixed—with money that appeared from a safety budget with no cap.

This is the reversal of what technology has done to the modern workplace. Where technology centralizes decisions in algorithms and dashboards, O’Neill decentralized decisions to the people with the most information. He broke up Alcoa’s command-and-control structure into twenty autonomous business units, each reporting directly to him.¹⁰ Information flowed up. Authority flowed down. The human being at the point of contact—not the system—was the decision-maker.

Mechanism Three: Shared Prosperity Replaces Shared Fear

O’Neill aligned everyone’s economic interest with the company’s performance. He devised a “floating dividend” for shareholders—higher dividends in exceptional years. Rank-and-file workers received profit bonuses. Executive pay was tied to profitability.¹¹ And he promised the unions that productivity gains would not cost jobs. He kept that promise: Alcoa’s workforce grew from 59,000 to 140,000 across 36 countries.¹²

Crucially, he did not pay people for safety. He paid them through the profitability that safety-driven improvements generated. O’Neill himself said: “I don’t think you should pay people for adherence to values. You should expect them to adhere to values.”¹³ The incentive was not a bounty for compliance. It was participation in the wealth that agency created. The difference is structural: compliance incentives reward obedience. Agency incentives reward contribution. Only contribution creates value.

5

What O’Neill set in motion was not a safety program. It was a value creation cascade—a chain reaction in which restoring personal agency at the bottom of the organization produced financial results at the top, each link reinforcing the next.

Here is how each link worked at Alcoa.

Personal Agency → Employee Satisfaction. When workers believed their observations mattered—when they could stop a line, call the CEO, and see the problem fixed—satisfaction rose. Not because of perks or compensation, but because of meaning. The worker was no longer a pair of hands executing a procedure. The worker was a partner in an enterprise. Turnover fell. Union relations improved. George Becker, president of the United Steelworkers, called O’Neill’s commitment “very sincere” and “one of his strong points.”¹⁴

Employee Satisfaction → Discretionary Effort → Process Improvement. Satisfied employees do not merely show up. They contribute. The safety-reporting infrastructure became a pipeline for every kind of operational idea—quality improvements, workflow suggestions, waste reduction. Because safety required examining every production process in detail, the improvements discovered along the way were enormous. One year after implementation, Alcoa hit a record in profits.¹⁵

Efficiency & Productivity → Quality → Customer Satisfaction. O’Neill argued that a process flawed enough to injure workers was too flawed to produce high-quality products efficiently.¹⁶ The inverse proved equally true: a process safe enough to protect workers was a process refined enough to deliver superior quality to customers. Manufacturing cycle times fell. Defect rates fell. Customers received better products faster, and they noticed.

Revenue Growth → EBITDA Improvement. Productivity gains dropped directly to the bottom line. Revenue grew from $9.8 billion to $16.3 billion. But more importantly, the margin structure improved because the gains came from operational excellence—not from pricing tricks or financial engineering. Net income grew sevenfold, from $200 million to $1.484 billion.¹⁷

Enterprise Value → Exit Attractiveness. Market capitalization rose from $3 billion to $27.53 billion—an 818 percent increase.¹⁸ The stock price went from a split-adjusted $5 to $40, outperforming Microsoft and Intel.¹⁹ Alcoa became the most profitable company in an industry where everyone else was losing money. For a PE sponsor, this is the exit thesis made flesh: a company that can demonstrate sustainable, culture-driven performance improvement commands a premium multiple because the acquirer is not buying a one-time cost reduction. The acquirer is buying a system that generates continuous improvement independent of any single leader—as proven by the fact that Alcoa’s safety and performance metrics continued improving for more than two decades after O’Neill left.²⁰

6

The modern workplace is moving in the opposite direction. Every year, more decisions are automated, more workers are monitored, more discretion is replaced by protocol. The instinct is understandable: technology is consistent, scalable, and auditable. But the cost is invisible and compounding. Each layer of automation that removes a human decision removes a human being’s reason to care. Each compliance system that monitors an employee’s behavior tells that employee, implicitly, that their judgment is not trusted. And each dashboard that centralizes information at headquarters takes it away from the person who could actually use it on the factory floor, in the customer call, at the loading dock.

O’Neill proved that the most powerful technology is not the one that replaces human judgment. It is the one that amplifies it. His real-time safety-reporting system did not monitor workers. It gave workers a voice that carried from a smelting floor in Tennessee to the CEO’s office in Pittsburgh in twenty-four hours. The technology was in service of agency. Agency was in service of value creation. And value creation was not a phase that followed the turnaround. It was the turnaround.

7

The lesson for PE sponsors is urgent, because the mistake is being made right now, in portfolio companies across every fund.

When a portco is in distress, the instinct is compliance: install the monitoring, enforce the reporting, layer on the technology. This is the equivalent of what O’Neill rejected in 1987. It would have produced a modest reduction in Alcoa’s injury rate. It would not have produced $24.5 billion in new enterprise value. It would not have grown EBITDA sevenfold. And it would not have made the company the most attractive acquisition target in its industry.

Compliance solutions protect against downside. They do not create upside. And in a distressed company, protecting against downside without creating upside is a slow liquidation with better documentation.

The alternative is what O’Neill proved: restore personal agency. Find the one shared interest that unites owners, operators, lenders, and employees. Build the incentive architecture for them to pursue it. Use technology to amplify their voice, not to replace their judgment. And trust that people who believe their contribution matters will produce results that no compliance regime, no monitoring system, and no algorithm can generate.

Turnaround and value creation are not sequential phases. They are symbiotic—two expressions of the same underlying force, which is the individual human being’s capacity to observe, decide, and act when given the agency to do so. O’Neill did not save Alcoa. He gave 59,000 people the tools and the permission to save it themselves.

That is the agency dividend. It compounds forever. And it is the only thing a buyer pays a premium for at exit.