The opioid epidemic did not happen because prescription painkillers are dangerous. It happened because one family built a machine — a machine that bribed physicians, corrupted a federal regulator, laundered their reputation through Harvard and the Louvre, moved their money offshore when investigators came, and then used the bankruptcy courts to buy their way to immunity while 500,000 Americans died. This is a map of that machine. Every connection is sourced to a primary document.
In January 1996, a memo circulated inside Purdue Pharma's marketing department. OxyContin had just received FDA approval. The company's sales leadership was preparing its launch strategy. The memo described the anticipated prescription volume in language that, two decades later, a federal judge would enter into evidence at a criminal sentencing hearing.
"It will be followed by a blizzard of prescriptions that will bury the competition. The prescription blizzard will be so deep, dense, and white."
The word "white" was not accidental. OxyContin tablets were white. But the people who wrote that sentence knew exactly what they meant. They were describing an avalanche. They were describing what would become the largest pharmaceutical fraud in American history — one that would kill approximately 500,000 people over the next two decades, [1] enrich the family that owned the company by an estimated $12 billion, [2] and ultimately end in a Supreme Court ruling that is still reshaping how American corporations use bankruptcy to escape civil liability.
This is not the story of a drug that turned out to be more dangerous than anyone expected. The internal documents released in litigation show clearly that Purdue Pharma knew, from the beginning, that OxyContin was highly addictive. They knew the abuse-deterrence claims they were making to physicians were not supported by clinical evidence. They knew their sales force was targeting the highest-prescribing physicians in the country regardless of whether those physicians were under investigation for diversion. They knew, and they kept going.
What made this possible was not just corporate greed. It was a machine — a carefully constructed network of paid physicians, corrupted regulators, laundered reputations, and political influence — that the Sackler family built over decades and used to insulate itself from accountability until the bodies made that insulation impossible to maintain.
To understand what Purdue Pharma was, you have to understand what Arthur Sackler was. Arthur was the oldest of the three Sackler brothers — Arthur, Mortimer, and Raymond — who together purchased Purdue Frederick, a small pharmaceutical company, in 1952. [3] Arthur did not run Purdue. His specialty was something different: he was one of the founding architects of modern pharmaceutical advertising.
In the 1950s and 1960s, Arthur Sackler ran a medical advertising agency called William Douglas McAdams, which produced campaigns for pharmaceutical companies including Pfizer and Roche. He is widely credited with inventing the practice of marketing prescription drugs directly to physicians — deploying sales representatives who visited doctors in their offices, creating a sense of peer endorsement and urgency around specific medications. [4] One of his campaigns, for the tranquilizer Valium, helped make it the most prescribed drug in America by the 1970s.
Arthur died in 1987, eight years before OxyContin was approved. His estate sold his one-third option in Purdue Frederick to his brothers Mortimer and Raymond. [5] But the playbook Arthur built — the physician-targeting, the sales force model, the cultivation of institutional credibility through donations — that playbook did not die with him. His brothers applied it to OxyContin, and they applied it with an intensity that Arthur never attempted with Valium.
The Sackler family's cultural philanthropy — the museum wings, the university buildings, the endowed chairs — was not generosity. It was strategy. It was the same strategy Arthur had pioneered in pharmaceutical advertising: the strategy of institutional credibility. If your name is on the Louvre, you are not a drug dealer. You are a patron of civilization.
On December 12, 1995, the FDA approved OxyContin — a controlled-release formulation of oxycodone — for the treatment of moderate to severe pain. [6] The approval was handled by a medical officer in the FDA's Division of Anesthetic, Critical Care and Addiction Drug Products named Dr. Curtis Wright.
The label Wright approved contained a sentence that Purdue Pharma would quote in its sales materials, its physician education programs, and its congressional testimony for the next seven years: that OxyContin's "delayed absorption" was believed to reduce the potential for abuse. [7] That claim had not been demonstrated in clinical trials. It was, in the language of the 2007 federal guilty plea, a misrepresentation — a claim the company made to physicians knowing it was not supported by the evidence.
Nineteen months after approving OxyContin, Curtis Wright left the FDA and took a job at Purdue Pharma. [8] He was eventually paid more than $400,000 per year. [9] The FDA's conflict-of-interest rules at the time imposed a two-year cooling-off period before a regulator could work on matters that came before his former division. They did not prohibit him from working at the company whose flagship product he had personally approved.
"The label that Dr. Wright approved was used by Purdue's sales force as a marketing claim for years. The FDA's own records show it was not supported by clinical evidence at the time of approval."
— GAO-04-110: Prescription Drugs — OxyContin Abuse and Diversion and Efforts to Address the Problem. U.S. Government Accountability Office, December 2003, pp. 23-26.The Wright revolving door was not a secret at the time — it was reported by congressional investigators in 2002 and flagged in the GAO report in 2003. It produced no regulatory action. It produced no criminal referral. It produced, instead, a debate about the adequacy of FDA conflict-of-interest rules that continued for more than a decade while OxyContin prescriptions climbed every year.
In 1980, the New England Journal of Medicine published a letter — not a study, a letter, five sentences long — from two Boston University researchers named Hershel Jick and Jane Porter. The letter reported that among nearly 12,000 hospitalized patients who had received narcotics for pain, only four had been documented as developing addiction. The letter concluded: "We conclude that despite widespread use of narcotic drugs in hospitals, the development of addiction is rare in medical patients with no history of addiction." [10]
This letter — five sentences, no methodology section, no follow-up study, written about patients receiving narcotics in a hospital setting for acute pain — became Purdue Pharma's primary scientific citation for the claim that OxyContin was not addictive. Sales representatives were trained to reference it. Physician education materials cited it. By 2001, the letter had been cited more than 600 times in medical literature, in most cases to support the claim that opioids prescribed for chronic pain carried a low risk of addiction. [11]
Jick himself, who is now in his eighties, has been interviewed repeatedly about the misuse of his letter. He told the New York Times in 2017 that he had no idea his letter would be used this way. "I'm essentially mortified that that letter to the editor was used as an excuse to do what these drug companies did," he said. [12]
The Porter-Jick letter was not the only scientific misrepresentation Purdue deployed. But it was the most consequential, because it gave the sales force something to hand to skeptical physicians — a peer-reviewed citation, however thin, that appeared to validate the company's core claim. A 2017 reanalysis published in the New England Journal of Medicine traced 608 citations of the original letter and found that the overwhelming majority cited it as evidence that addiction risk in non-hospitalized chronic pain patients was low — a conclusion the original letter explicitly did not reach and could not support. [13]
Richard Sackler, son of Raymond Sackler, joined Purdue Pharma's board in the 1990s and served as President from 1999 to 2003. [14] He was not a passive executive. The internal documents released by the Massachusetts Attorney General's office in 2019 — the first time the public had seen Purdue's internal communications — show a man personally obsessed with OxyContin's growth and personally involved in shaping how the company responded to evidence of addiction.
At OxyContin's launch party in 1996, Richard Sackler gave a speech. He did not speak about pain relief. He did not speak about the patients the drug was intended to help. He said, in the language of a sales campaign: [15]
"We have to be in the face of prescribers every day. We have to be more persistent than our competitors. We have to be willing to make more calls, see more doctors, and never take no for an answer."
As reports of addiction and abuse began accumulating in the early 2000s, Richard Sackler's private communications show a consistent pattern: dismiss the evidence, attack the credibility of those raising it, and redirect responsibility toward the people who had become addicted. In a February 2001 email — produced in litigation and quoted in the Massachusetts complaint — he wrote that the company's response to the growing addiction crisis should be to attack the problem from a different direction. He described people who had become addicted to OxyContin as "criminals" who were "the culprits and the problem." [16]
In 2015, Richard Sackler was deposed by attorneys representing states and counties suing Purdue. The deposition was sealed. A Kentucky court ordered it unsealed in January 2019. In the deposition, Sackler was asked repeatedly whether he bore any responsibility for the opioid epidemic. His answer, given dozens of times in different formulations, was no. He claimed he could not recall specific marketing decisions. He claimed he was not aware of studies showing OxyContin's high potential for abuse. He claimed he had acted at all times in good faith. [17]
Richard Sackler has never been criminally charged. He is estimated to be personally worth approximately $3 billion.
From 2004 through at least 2019 — a fifteen-year span that covered the peak of the opioid crisis, a federal guilty plea, and thousands of civil lawsuits — McKinsey & Company provided management consulting services to Purdue Pharma. [18] This was not passive work. The internal McKinsey documents produced in the federal opioid multidistrict litigation show that consultants actively advised Purdue on how to sustain and increase OxyContin sales during the period when the company was under the most intense public and legal scrutiny it had yet faced.
One McKinsey slide deck from 2017 — the year the opioid crisis was declared a national public health emergency — proposed a mechanism for incentivizing pharmacy chains to continue dispensing OxyContin despite mounting diversion concerns. The proposal involved paying the pharmacy companies a rebate for each OxyContin overdose that occurred among patients who had filled a prescription at their stores. The logic was explicit: if pharmacies bore a financial cost for each overdose, they could be compensated for that cost, and thereby neutralized as a source of resistance to continued distribution. [19]
"McKinsey advised Purdue on how to 'turbocharge' sales to high-risk patients — including patients whose doctors were already under investigation for overprescribing."
— Massachusetts Attorney General civil complaint against McKinsey, Feb. 2021. Source: Mass. AG press release, Feb. 4, 2021.A separate McKinsey analysis identified Purdue's highest-value prescribers — physicians writing the most OxyContin prescriptions — and recommended that the sales force focus resources on those physicians. The analysis did not distinguish between high-prescribing physicians who were appropriately managing complex pain patients and high-prescribing physicians who were operating pill mills under active DEA investigation. The criterion was volume. [20]
In February 2021, McKinsey agreed to pay $573 million to settle claims brought by 47 states, the District of Columbia, and five U.S. territories. [21] McKinsey did not admit wrongdoing. The firm's senior partner who led the Purdue engagement, Martin Elling, resigned from McKinsey in the days before the settlement was announced. No McKinsey employee has been criminally charged in connection with the firm's Purdue work.
While Purdue Pharma's sales force was executing the blizzard, the Sackler family was engaged in a parallel project: the systematic purchase of cultural credibility at the highest institutions in the Western world. The scale of this effort — and the degree to which it insulated the family from scrutiny for more than two decades — is documented in extraordinary detail by the institutions themselves, which were eventually forced to explain, publicly, why they had taken the money and why they were reluctant to give it back.
The list of institutions that accepted Sackler donations and attached the family name to buildings, galleries, and endowed chairs spans five countries and four decades. [22] In the United States alone, it includes: Harvard University (two buildings — the Sackler Building and the Arthur M. Sackler Museum, which houses Harvard's collection of Asian and Islamic art); the Metropolitan Museum of Art in New York (the Sackler Wing, which contains the Temple of Dendur); the Solomon R. Guggenheim Museum; the American Museum of Natural History; the Smithsonian Institution (the Arthur M. Sackler Gallery on the National Mall). [23]
In Europe: the Louvre in Paris (the Sackler Wing of Oriental Antiquities, named since 1997 and removed July 2019); the British Museum in London (the Raymond and Beverly Sackler Rooms, renamed 2022); the Tate galleries (removed 2022); the Victoria and Albert Museum; the Serpentine Galleries; the National Portrait Gallery; the Ashmolean Museum at Oxford; the Sackler Library at Oxford (Oxford severed ties with the family in 2023). [24]
These were not arms-length transactions. The Sackler family did not simply write checks to worthy institutions. They negotiated naming rights. They secured board positions. They cultivated relationships with university presidents, museum directors, and curators who would in turn lend their institutions' credibility to events and publications that further elevated the Sackler name. When Harvard President Lawrence Bacow was asked why the university would not return the Sackler family's donations and rename the buildings, he cited "legal and contractual obligations." [25] This was true. The contracts had been negotiated by lawyers. They had been written specifically to make it very difficult to un-name a building.
The settlements that the Sackler family eventually reached with their various creditors contained a provision that, in the context of the broader story, is remarkable. The terms of the 2022 revised bankruptcy settlement allowed museums and universities that had received Sackler money to remove the family name from their buildings — but only if they provided 45 days' confidential notice and agreed not to "disparage the Sacklers." [26] Even in their settlement documents, the Sackler family's lawyers were still controlling the narrative.
In 2017, photographer Nan Goldin was prescribed OxyContin after a wrist surgery. She became addicted. She entered treatment. When she came out, she read a New Yorker article about Purdue Pharma and the Sackler family and she recognized the name — from museums, from galleries, from the institutions she had spent her career working within. [27]
She founded an activist group called P.A.I.N. — Prescription Addiction Intervention Now. The group's tactic was simple and, it turned out, highly effective: stage protests inside the cultural institutions that bore the Sackler name, filling their most famous spaces with empty prescription bottles and red-stained fake opioid dollar bills, making the connection between the family's philanthropy and the epidemic impossible to ignore. In January 2018, P.A.I.N. protested at the Guggenheim, scattering hundreds of fake prescription slips from the museum's famous rotunda down to the atrium floor. In 2018, they marched from Harvard Square to the Arthur M. Sackler Museum. In 2019, Goldin led a protest inside the Louvre, which removed the Sackler name from its walls within days. [28]
The institutional response to P.A.I.N.'s campaigns was a study in the tension between stated values and financial dependency. The Tate, the Guggenheim, and the National Portrait Gallery all declined to accept further Sackler donations within weeks of the first protests. But declining future donations was the easy move — it cost nothing. Removing existing names from existing buildings, returning existing donations, acknowledging what the money was — that was harder, and in most cases, it did not happen until years later, and only because a court settlement made it possible.
Goldin's own statement about why she targeted the museums is one of the most precise descriptions of the Sackler strategy that exists: "I believe the Sacklers live in their museums. That's what they care about. So that seemed the place to get their ear." [29] She was right. Within three years of the first protests, nearly every major institution that had accepted Sackler money had either removed the name, declined future donations, or begun formal proceedings to do both.
Between 2008 and 2018 — as the opioid death toll mounted, as state attorneys general began filing suit, as congressional investigations were launched, and as DEA scrutiny of Purdue's distribution practices intensified — the Sackler family withdrew an estimated $10.7 billion from Purdue Pharma. [30]
This is the number that defines the moral scale of what happened. The family did not withdraw this money before the crisis was apparent. They withdrew it during the crisis. They withdrew it while their lawyers were arguing in courts across the country that Purdue had done nothing wrong. They withdrew it, in significant part, through a network of trusts and offshore accounts designed to place the money beyond the reach of American courts. [31]
The bankruptcy examiner's report, filed in 2021, traced the specific mechanisms by which the money moved. Wire transfers to accounts in Switzerland. Distributions to trusts established in jurisdictions with strong creditor protection. Transfers to family members who were not themselves parties to any litigation and who would later argue, in the Supreme Court proceedings, that they could not be held liable for the conduct of a company they did not personally control. [32]
"The Sackler family was withdrawing hundreds of millions of dollars a year from Purdue even as litigation was mounting. The timing was not coincidental."
— New York Attorney General Letitia James, press statement, March 2022. Source: NY AG press release, March 3, 2022.The precise figure is contested. The Sackler family's attorneys argue that distributions to family members were legitimate dividends from a profitable company and that the bankruptcy examiner's methodology overstates what was "extracted" versus what was legitimately earned. The House Oversight Committee's figure of $10-12 billion comes from the committee's own analysis of Purdue's financial records. The bankruptcy examiner's independently derived figure was $10.7 billion. [33] The three figures converge on the same conclusion: the family systematically removed wealth from the company, at scale, during the period when that company's liability was becoming clear.
On September 15, 2019, Purdue Pharma LP filed for Chapter 11 bankruptcy protection in the Southern District of New York. At the time of the filing, the company faced more than 2,600 lawsuits from states, cities, counties, Native American tribes, hospitals, and individuals. [34]
The bankruptcy filing was not, primarily, a financial necessity. Purdue was not insolvent in the conventional sense. It was a strategic maneuver: by filing for bankruptcy, Purdue could consolidate all pending litigation into a single proceeding, eliminate the possibility of a runaway jury verdict in any individual state, and negotiate a global settlement under conditions that it could substantially control. This use of Chapter 11 as a litigation shield — available to solvent corporations facing mass tort claims — has been controversial for decades. The Purdue case became its most prominent test. [35]
The most controversial element of the resulting settlement plan was not the dollar amount. It was the immunity. The Sackler family — whose members had not themselves filed for bankruptcy and who had, as described above, moved billions of dollars out of reach of American courts in the years preceding the filing — sought, as a condition of their $6 billion contribution, complete and permanent immunity from any current or future civil lawsuit arising from OxyContin. This immunity would cover not just Sackler family members who had been named in existing suits but all Sackler family members, including those against whom no lawsuit had ever been filed. [36]
In September 2021, the bankruptcy judge presiding over the case, Judge Robert Drain of the S.D.N.Y., confirmed the plan and granted the immunity. The Department of Justice appealed. The District Court reversed. The Second Circuit reinstated it. And then, on June 27, 2024, the Supreme Court ruled 5-4 in Harrington v. Purdue Pharma LP that the bankruptcy courts had exceeded their statutory authority. [37] The Sackler family immunity was struck down. A revised settlement was ordered.
In the crowded creditor landscape of the Purdue bankruptcy — states, counties, cities, Native American tribes, hospitals, insurance companies, all with lawyers and all fighting for priority — there was one attorney whose specific mission was to represent the people who had actually been harmed: the individuals and their families. Her name is Anne Andrews, and she is the managing partner of Andrews & Thornton, based in Los Angeles.
Andrews was selected by the United States Trustee for the Southern District of New York to serve in a representative capacity on the Official Committee of Unsecured Creditors in the Purdue bankruptcy — one of the most complex and consequential bankruptcy proceedings in American legal history. [38] Through that committee and through direct representation, her firm represented approximately 60,000 individual claimants — people who had become addicted, people who had lost family members, veterans whose injuries had been treated with Purdue opioids and who had developed dependence. [39]
The structural problem Andrews faced was this: in a bankruptcy proceeding with hundreds of billions of dollars in claimed liability and a finite pool of available assets, the entities with the most organized and best-funded legal representation — state governments, counties with dedicated legal departments, major hospital systems — would, by default, receive the majority of whatever was distributed. Individual victims, represented piecemeal or not at all, would be last in line. The initial settlement plan reflected this: approximately $750 million was designated for individual victims out of an $11 billion total plan. [40]
"With governmental claims tallying in the trillions of dollars, it was an uphill battle from day one for individual victims to have any recovery guaranteed. Anne Andrews went toe-to-toe with lawyers for hundreds of cities and counties to ensure that opioid victims would receive some recovery."
— Andrews & Higgins firm statement on the Purdue Pharma bankruptcy. Source: andrewshiggins.com, 2024.Andrews fought that allocation through multiple rounds of litigation, appeals, and negotiation. When the revised settlement was confirmed following the Supreme Court's 2024 ruling, individual victims received $850 million — a number that, in the context of $12 billion extracted by the Sackler family, remains a fraction, but a fraction that exists because one attorney refused to allow the individuals to be eclipsed entirely by the institutional creditors. The bankruptcy court, in confirming the revised plan, specifically acknowledged Andrews and her partner Sean Higgins as instrumental in securing the individual victim allocation. [41]
Andrews brought to the Purdue case a track record built specifically for this kind of fight. She had previously served as chair of the unsecured creditors' committee in the Insys Therapeutics bankruptcy — the first opioid manufacturer bankruptcy — and had secured precedent-setting recoveries for individual victims in that case. She had represented victims in the Mallinckrodt opioid bankruptcy and in the PG&E wildfire bankruptcy. The lessons from each case were applied to the next. [42]
Approximately 500,000 Americans died of opioid overdoses between 1999 and 2019. [43] That figure does not include the people who became addicted and survived, the families destroyed, the children placed in foster care, the communities hollowed out, the criminal justice resources consumed, or the healthcare costs incurred. A 2021 study estimated the total economic cost of the opioid epidemic between 2002 and 2017 at approximately $1.5 trillion. [44]
The Sackler family extracted approximately $12 billion from Purdue Pharma during the period the crisis was ongoing. They agreed to contribute $6 billion to settle all claims — a figure representing half of what they took out. The individual victims who filed claims in the bankruptcy will receive approximately $850 million in total — less than one-tenth of one percent of the $1.5 trillion in estimated economic damage, and about 7 cents on every dollar the family extracted. [45]
No member of the Sackler family has been criminally charged. The company they owned pleaded guilty to federal crimes in 2007 and again in 2020. Three executives pleaded guilty to misdemeanor charges in 2007 — not felonies, misdemeanors — and none served a single day in prison. [46] The consultants who advised them on how to keep selling during the crisis paid a civil settlement and admitted nothing.
The machine the Sackler family built — the sales force, the corrupted regulator, the captured physicians, the weaponized science, the cultural institutions, the political donations, the offshore trusts, the bankruptcy court gambit — performed almost exactly as designed. It transferred the cost of what they did onto the rest of America, while the family kept most of the money.
The 500,000 death figure requires a precision the original case against Purdue does not always supply. Credentialed public health researchers — including Keith Humphreys of Stanford and David Herzberg of SUNY Buffalo, author of White Market Drugs (2020) — have published peer-reviewed work making a distinction the bankruptcy proceedings tended to obscure: the opioid crisis has two phases with different primary causes. [47]
The first phase, roughly 1996 to 2010, is the one this case file documents: prescription opioids, driven by Purdue's marketing, physician capture, and regulatory failure. The death toll in this phase is directly traceable to OxyContin and the market it created. The second phase, from approximately 2013 onward, is driven primarily by illicit fentanyl — a synthetic opioid 50 to 100 times more potent than morphine, manufactured outside the pharmaceutical supply chain and smuggled across borders. [48] The post-2013 death curve does not track OxyContin prescriptions, which were falling after the 2010 reformulation. It tracks illicit supply.
This is not an exoneration of Purdue. The prescription wave created the population of dependent patients who, when prescriptions were cut off, turned to heroin and then fentanyl. The chain of causation runs from OxyContin to the illicit market. But the academic debate exists and is legitimate: Purdue's direct culpability for deaths after 2013 is a more contested evidentiary question than its culpability for deaths before 2010. The $850M settlement figure, the $21B distributor settlement, and the $8B DOJ penalty are all built on that contested second phase as well as the clear first phase. A complete investigation acknowledges the distinction.
There is a second contested question this case file would be incomplete without acknowledging. Before the OxyContin era, credentialed pain specialists argued — in peer-reviewed literature — that chronic pain was catastrophically undertreated in American medicine, and that opioid-based treatment was being withheld from patients who genuinely needed it. The American Academy of Pain Medicine's 1997 consensus statement, and a body of literature supporting it, made the case for broader opioid prescribing on humanitarian grounds. [49]
Those arguments were later weaponized by Purdue's sales force and used to justify reckless prescribing. But they did not originate with Purdue, and the physicians who accepted them were not all corrupt — many were acting on what their continuing medical education told them was good medicine. That CME, as this map documents, was itself contaminated by pharmaceutical funding. The structural failure was not simply that doctors were bribed. It was that the information environment doctors operated in had been systematically shaped by the industry whose products they were prescribing. The machine did not need corrupt doctors. It needed trusting ones.
Between 2006 and 2014, three pharmaceutical distributors — McKesson, AmerisourceBergen, and Cardinal Health — shipped 76 billion oxycodone and hydrocodone pills across the United States. [50] The DEA's Suspicious Order Monitoring system received data on these shipments in real time. The DEA did not stop them. This failure is documented in the House Energy and Commerce Committee's 2019 subcommittee investigation. The distributors settled for $21 billion in 2022. No DEA official has been charged or disciplined. The structural incentive — revolving door between DEA enforcement leadership and the distribution industry — is documented in public lobbying records and has not been adequately investigated.