Purdue's Crisis Response Wasn't a Failure. It Worked for Seventeen Years.

On May 10, 2007, Purdue Pharma stood in a federal courthouse in western Virginia and pleaded guilty to a crime. The company and three of its top executives admitted to criminal misbranding and agreed to pay $634,515,475.¹ By the standards of a corporate scandal, this looks like the end of the story: caught, charged, punished. It was not the end. It was the opening move. Because over the next ten years - May 2007 through at least March 2017 - Purdue kept doing a version of the very thing it had just pleaded guilty to, and the Justice Department would eventually prove it.³

The official story is that Purdue handled its crisis badly: it denied, it stonewalled, its public relations failed. The real story is the opposite. Purdue handled its crisis with cold competence. The denial was not a communications stumble; it was a strategy. And it worked - for seventeen years.

Here is the thesis a smart friend can repeat at dinner: Purdue's crisis response was a legal-and-financial maneuver, not a moral reckoning. Plead small. Keep selling. Pull the family's money out before the lawsuits could reach it. Then use bankruptcy to buy permanent immunity. Each step was rational. Each step bought time. The plan failed at exactly one point - and only because of who was holding the gavel.

The 2007 plea bought silence, not a confession

Look closely at what the company actually admitted in 2007 and what it did not. The crime centered on a single marketing claim, traceable to the drug's own 1996 label, which read that OxyContin's delayed absorption 'is believed to reduce the abuse liability of a drug.' Sales reps were trained to spin that hedged sentence into a hard promise: addiction risk 'less than one percent.'⁶ That was the misbranding. But the structure of the plea mattered more than the charge. The corporate entity took the felony-grade misbranding hit; the three executives pleaded to misdemeanors as individuals.¹ No one went to prison. The company wrote a check that, set against the money the drug was generating, functioned less like a punishment and more like a cost of doing business. Prescriptions had grown from roughly 300,000 in 1996 to nearly 6 million by 2001 - and OxyContin would go on to total more than $35 billion in sales.⁷ A $634 million settlement against that backdrop is not a deterrent. It is a toll.

And the decisive fact about 2007 is what came after it. Most people read a guilty plea as a turning point - the moment a company chastens and changes. Purdue's plea changed almost nothing operationally. The 2020 federal felony case lays it out: Purdue conspired to defraud the United States, including the Drug Enforcement Administration, from May 2007 through at least March 2017.³ The misconduct did not stop at the courthouse door. It continued, on the clock, the day after the company said it was sorry.

While the lawyers settled, the family was draining the tank

The most revealing part of the response isn't in the courtroom at all. It's in the cash flow. As litigation mounted and the family came to understand that the suits might reach them personally - not just the corporate shell - the Sacklers ran what the Supreme Court's own opinion calls a 'milking program': they withdrew roughly $11 billion from Purdue, about 75% of the company's total assets, over the decade before bankruptcy.⁴ Read that mechanism slowly, because it is the whole game. A company facing ruinous liability is worth pursuing only to the extent it has money. Move the money out of the company and into the family, and you have hollowed out the very target every plaintiff is aiming at. By the time Purdue filed for Chapter 11 in 2019, the thing creditors could seize was a husk; the wealth that the husk had generated sat somewhere claims could not easily follow.⁴

This is the difference between a crisis you survive by accident and one you manage on purpose. The timing of the withdrawals tracks the rising litigation risk, not ordinary dividends. The internal record reinforces how deliberate the underlying machine was from the start: unsealed Purdue documents include a Sept. 3, 1996 email from Richard Sackler noting first-year sales goals were hit four months early, and emails documenting a strategy to push the drug at non-cancer chronic-pain patients.⁸ The same intentionality that built the sales engine ran the escape.

The last move was the boldest: buying immunity that was never for sale

The plan's capstone was the most audacious step. The Sacklers themselves never filed for bankruptcy - the company did. Yet the proposed reorganization would have used Purdue's bankruptcy to extinguish opioid-related claims against the family, in exchange for the Sacklers returning $4.325 billion.⁴ Strip away the legalese and the trade is stark: pay back a fraction of what you withdrew, and in return receive a court order wiping out everyone's right to sue you - whether or not those people agreed. That is the entire design. Withdraw $11 billion, offer back roughly $4 billion, and have a judge declare the matter closed against people who never consented and were never themselves in bankruptcy.⁴ If it had held, it would have been the cleanest corporate escape in modern memory: a family kept the bulk of the wealth and bought blanket peace with a slice of it.

It did not hold. On June 27, 2024, the Supreme Court ruled 5-4 in Harrington v. Purdue Pharma that the Bankruptcy Code does not authorize a release that discharges claims against a non-debtor - the Sacklers - without the consent of the people holding those claims.⁵ The single mechanism the entire plan depended on, the non-consensual third-party release, was the one thing the law would not stretch to provide. The playbook had survived a guilty plea, a decade of continued misconduct, an $11 billion drain, and a bankruptcy filing. It died on a question of statutory authority by a one-vote margin.

Isn't this just hindsight dressed as strategy?

The fair objection is that calling this a 'playbook' imposes a tidy design on what may have been a series of separate, defensive, lawyer-driven reactions - each settlement and withdrawal made for its own reasons, only looking coordinated in the rearview mirror. That deserves a real answer. Two facts make the coincidence hard to credit. First, the misconduct demonstrably continued for a full decade after the 2007 plea, which rules out the most charitable reading - that 2007 was a genuine attempt to reform that simply fell short.³ A company that meant to change does not keep conspiring to defraud the DEA. Second, the Court itself characterized the withdrawals as a 'milking program' driven by fear of personal liability - not as ordinary distributions, but as a deliberate response to litigation risk.⁴ You don't need a single secret memo titled 'The Plan' to call something a strategy. You need a consistent pattern of choices that all point the same direction: keep the revenue, move the wealth, buy the immunity. Every documented step points there. The honest concession is that no one move was illegal on its face - which is precisely what made the response so effective for so long.

Purdue handled its defining crisis the way a skilled litigant handles a losing case: it never tried to win the argument, only to control the cost. Plead to the smallest charge, keep the revenue flowing, drain the wealth ahead of the claims, and then ask a court to seal the exit. For seventeen years it nearly worked. The thing that finally stopped it was not public outrage, not the size of the harm, not even a confession - the company had already confessed twice. It was a five-vote majority deciding that one specific door the law had never opened would stay shut. The most expensive crisis response in pharmaceutical history failed on a technicality of bankruptcy authority - which tells you exactly how close a well-managed escape can come to working.