The fraud that ended the telecommunications boom was discovered at night, by accident, by a woman who had been told to stop looking. In the late spring of 2002, Cynthia Cooper, the head of internal audit at WorldCom, began pulling accounting entries after hours from the company’s systems in Clinton, Mississippi, working quietly with a small team because she suspected — correctly — that if the finance department knew what she was doing, the files might become harder to find. What she and her colleagues eventually located, buried in the ledgers of the second-largest long-distance carrier in America, was $3.8 billion in expenses that had been converted, by journal entry, into assets. The number would keep growing. By the time the restatements were finished, WorldCom had overstated its financial results by more than $11 billion, the largest accounting fraud in American history to that point — larger than Enron, the scandal it displaced within months of Enron’s own collapse.

The Coffee Shop

Every account of Bernard John Ebbers begins with the improbability of him. He was born in Edmonton, Alberta, in 1941, one of five children, and his early résumé reads like a parody of the self-made man: milkman, bouncer, delivery driver, high-school basketball coach. He came south to Mississippi College on a basketball scholarship, stayed, and went into the motel business, assembling a small chain of Best Westerns across Mississippi and Texas. He was six feet four, wore cowboy boots and jeans to board meetings, taught Sunday school at Easthaven Baptist Church in Brookhaven, and led company meetings in prayer. Wall Street analysts, who ordinarily punish eccentricity, decided in his case that the folksiness was the point. Here was a man who seemed incapable of financial sophistication, which made the sophistication of his acquisitions seem like genius.

The company that became WorldCom was conceived in 1983 in a coffee shop in Hattiesburg, Mississippi, where a group of local businessmen sketched out a plan to resell long-distance capacity in the deregulated aftermath of the AT&T breakup. The firm was called LDDS — Long Distance Discount Services — and it was, at first, a failure. Ebbers, an early investor, took over as chief executive in 1985 and applied the only strategy he ever really had: buy. LDDS acquired small carriers across the South, then bigger ones, then bigger ones still. In 1995 the company renamed itself WorldCom. In 1996 it bought MFS Communications, and with it UUNET, one of the backbones of the early commercial Internet, for roughly $12 billion. And in 1998 it swallowed MCI — a company more than twice its size — in a deal valued at around $37 billion, at the time the largest merger in American corporate history.

By 1999, WorldCom’s stock had touched sixty-four dollars a share and the company was worth more than $180 billion. Ebbers had made some sixty acquisitions in fifteen years. The business press called him the Telecom Cowboy. What almost nobody outside the company understood was that the serial acquisitions were not just a growth strategy; they were an accounting strategy. Each deal brought a fresh round of merger reserves and write-offs that could be released into earnings later, a way of smoothing results that depended on the next deal always arriving. In October 1999, Ebbers announced the deal that would have kept the machine running — a merger with Sprint valued at well over a hundred billion dollars. In the summer of 2000, American and European regulators killed it. The music stopped. The industry was drowning in fiber-optic overcapacity, long-distance prices were collapsing, and WorldCom, for the first time in its existence, had to be an operating company rather than an acquiring one.

The Line-Cost Problem

The fraud, when it came, was almost insultingly simple. WorldCom’s largest expense was what the industry calls line costs — the fees paid to other carriers to complete calls and lease network capacity. As revenue flattened, line costs consumed a growing share of every dollar, and the ratio that analysts watched most closely began to deteriorate. Beginning in 2000, under the direction of chief financial officer Scott Sullivan, the company simply moved billions of dollars of line costs off the expense ledger and onto the balance sheet, reclassifying ordinary operating costs as capital investments that could be depreciated over years rather than recognized immediately. Earlier, the company had drained reserve accounts to inflate results. Together, the adjustments manufactured billions in phantom profit. In quarters when WorldCom told the public it was comfortably profitable, it was in fact losing money.

Sullivan, who had been named the finest CFO in his industry by CFO magazine, did not do the bookkeeping himself. The entries were made by mid-level accountants in Clinton — the controller David Myers, the accounting director Buford Yates, and staff accountants Betty Vinson and Troy Normand — who understood that what they were being asked to do had no support in accounting principles, said so, and did it anyway. Vinson, by her own later account, wanted to resign and could not afford to; she had a family, and Clinton, Mississippi, was not thick with jobs for accountants. The moral architecture of the WorldCom fraud is, in this sense, a textbook — literally; it is taught in auditing courses — of how ordinary professionals are conscripted into extraordinary crimes one journal entry at a time: an instruction from above, an assurance that it is temporary, a fear of the consequences of refusing, and a quarter-by-quarter descent in which each entry makes the last one impossible to confess.

The external auditor, Arthur Andersen, certified WorldCom’s statements throughout — the same firm whose Houston office was, in the same period, shredding Enron documents. Andersen’s work papers for WorldCom rated the engagement a maximum-risk client, and yet its procedures never caught the reclassification of billions of dollars of line costs, in part because the company restricted the auditors’ access to the general ledger. The SEC, for its part, had begun asking questions in early 2002, prompted by the company’s eroding stock and a series of aggressive disclosures. But the decisive act of detection came from inside: Cooper’s team, tipped by anxieties percolating through the finance department, traced the capitalized costs entry by entry. When Cooper confronted Sullivan, he asked her to delay the audit. She declined. On June 25, 2002, WorldCom announced that it had inflated its results by $3.8 billion, and the board fired Sullivan. Less than a month later, on July 21, the company filed the largest bankruptcy in American history.

The Loans

Ebbers was already gone by then, and the manner of his going tells its own story about the American board of directors circa 2002. Ebbers had borrowed colossally against his WorldCom stock to finance a private empire — a half-million acres of timberland, a ranch in British Columbia said to be among the largest in Canada, a yacht-building yard, a soybean operation. When the stock fell, the margin calls came, and rather than let its founder dump shares into a falling market, WorldCom’s board lent him the money — loans and guarantees that ultimately exceeded four hundred million dollars, extended at rates far below what any bank would have charged, to cover the personal leverage of the chief executive. The directors approved it with barely a murmur. In April 2002, with the stock in single digits and the SEC circling, the board finally forced Ebbers out. He drove away from headquarters with a severance package that promised him $1.5 million a year for life — a promise the bankruptcy extinguished.

On the Sunday after his resignation, Ebbers stood before his congregation at Easthaven Baptist and told them, “I just want you to know you aren’t going to church with a crook.” It was the first public statement of what would become his entire legal defense: that whatever had happened inside WorldCom’s ledgers, he had not known about it. He was a coach, a deal man, a motel operator who read the Bible and not the balance sheet. The claim had a surface plausibility — Ebbers was famously indifferent to technology, reportedly avoided e-mail, and cultivated his innumeracy the way other executives cultivate golf handicaps. The question a jury would eventually have to answer was whether the cultivated ignorance of a chief executive is a defense or a technique.

The Aw-Shucks Defense

The government’s case, brought by federal prosecutors in the Southern District of New York, depended almost entirely on Scott Sullivan. The CFO pleaded guilty in 2004 to securities fraud and agreed to testify against the only man above him in the hierarchy. On the stand, Sullivan told the jury that he had repeatedly warned Ebbers that the adjustments being made to hit Wall Street’s numbers were improper, and that Ebbers had told him, in substance, that the company had to make its numbers. There were no e-mails from Ebbers directing the fraud, no memos, no recordings — the case was Sullivan’s word, corroborated by circumstance: Ebbers’s obsession with the stock price, his hundreds of millions in margin debt that made the stock price a personal emergency, his reputation inside the company as a micromanager who questioned the cost of coffee filters and demanded that headcount and capital budgets cross his desk.

Ebbers took the stand in his own defense — a gamble his lawyers reportedly resisted — and performed the innocence of a man betrayed by his subordinates. “I know what I don’t know,” he told the jury. He did not know technology. He did not know accounting. The cross-examination wrote itself: here was a man who had built a hundred-eighty-billion-dollar company through sixty acquisitions, negotiated the largest merger in American history, and presented himself to shareholders for fifteen years as the master of his business, now asking twelve citizens to believe that the business had been a mystery to him. On March 15, 2005, the jury convicted him on all nine counts — securities fraud, conspiracy, and seven counts of filing false statements with regulators.

On July 13, 2005, Judge Barbara S. Jones sentenced Ebbers, then sixty-three, to twenty-five years in federal prison — effectively a life sentence, and at the time the harshest punishment ever imposed on an executive of a major American corporation. The Second Circuit affirmed the conviction the following year, calling the sentence harsh but not unreasonable given the scale of the destruction: eleven billion dollars in fraudulent accounting, tens of billions in market value erased, seventeen thousand WorldCom workers who lost their jobs, and retirement accounts across the country — including state pension funds that had loaded up on a stock their consultants called safe — gutted. In September 2006, Ebbers drove himself, in his own Mercedes, to the federal prison at Oakdale, Louisiana, to begin serving his term.

The Ledger of Consequences

The people below Ebbers received the discounts that cooperation buys. Sullivan, the architect of the entries, was sentenced to five years. Myers and Yates received a year and a day each. Vinson got five months; Normand, probation. Cynthia Cooper, who had done the thing the entire apparatus of American financial gatekeeping had failed to do, was named one of Time’s Persons of the Year for 2002, alongside Enron’s Sherron Watkins and the FBI’s Coleen Rowley — three women who told the truth about institutions run by men. She later wrote that colleagues treated her not as a hero but as the person who had blown up the company, an experience so common among whistle-blowers that it amounts to a second finding of the scandal: the professional culture that produced the fraud also punished its detection.

The structural consequences were larger than any sentence. WorldCom’s collapse, arriving eight months after Enron’s, ended whatever argument remained against federal intervention in corporate accounting. Nine days after the bankruptcy filing, Congress passed the Sarbanes-Oxley Act, which for the first time required chief executives to personally certify their companies’ financial statements under criminal penalty — a provision aimed directly at the defense Ebbers would later attempt. Never again, the statute said in effect, would “I didn’t know” be an available posture for the person at the top. The board of directors that had lent Ebbers four hundred million dollars became a case study in governance failure; in a nearly unprecedented settlement, twelve former WorldCom directors agreed to pay roughly twenty-five million dollars out of their own pockets to settle investor litigation — personal payments, not insurance money, a signal to every boardroom in America that the job was not honorary.

Nor was the failure confined to auditors and boards. WorldCom’s rise had been championed on Wall Street by Jack Grubman, the Salomon Smith Barney telecommunications analyst who was, for most of the nineteen-nineties, the most powerful voice in the sector — and who maintained his enthusiasm for the stock nearly to the end, downgrading it only in the spring of 2002, when it traded in single digits and the damage to the investors who had followed him was complete. Grubman’s conflicts became a scandal of their own: he attended WorldCom board meetings, his bank collected enormous investment-banking fees from the company he covered, and congressional investigators established that WorldCom executives, including Ebbers, had received allocations of hot initial public offerings from Salomon — shares worth millions, distributed to the executives of the bank’s best client. In 2003, as part of the global research settlement that remade Wall Street’s rules, Grubman paid fifteen million dollars and accepted a lifetime ban from the securities industry, without admitting wrongdoing. The WorldCom fraud thus implicated, in a single case, every profession the market relies upon to check a chief executive: the CFO who engineered it, the accountants who booked it, the auditor who blessed it, the analyst who promoted it, and the directors who financed the boss’s margin calls while it happened. Each had a duty that, honored by any one of them, would likely have stopped the scheme years and billions earlier.

There is a school of thought, pressed by Ebbers’s defenders and by some students of the era, that he was the least culpable of the great scandal CEOs — that he never sold at the top (he was, in fact, ruined; his fortune went to creditors and the victim-compensation fund), that the fraud was Sullivan’s craft, and that twenty-five years for a first offense by a sixty-three-year-old man exceeded the sentences of murderers. The counterargument is the one the jury implicitly accepted: that the chief executive of a public company holds a professional trust, that the entire premise of the securities markets is that the numbers signed and certified to the public are real, and that a CEO whose personal solvency depends on the stock price, and who tells his CFO to hit the numbers, has committed the fraud whether or not he can name the journal entries. Ignorance, at that altitude, is not an accident. It is a design choice.

Ebbers served twelve years. In December 2019, Judge Valerie Caproni granted him compassionate release; he was seventy-eight, suffering from severe heart disease and dementia, blind in the functional sense, and had lost more than fifty pounds. His lawyers said he could no longer reliably recognize his family. He went home to Mississippi and died five weeks later, on February 2, 2020. The obituaries all reached for the same arc — milkman to mogul to inmate — and most of them quoted the Sunday-school line. What fewer noted was the detail that best captures the era he embodied: for years after the collapse, WorldCom’s successor, MCI, kept operating — the network was real, the calls went through, the business underneath the lie had substance. It was the numbers that were fiction, and the numbers were the only thing Wall Street had ever asked to see.