Institutional complicity

Institutional complicity describes how organizations can help sustain sexual harassment without every employee knowing what is happening or every decision-maker openly endorsing it. In the account of Harvey Weinstein, it took shape through legal agreements that muted claims, a company apparatus used to manage or contain allegations, and oversight that repeatedly treated financial exposure as a more urgent problem than women’s safety. Its force lay not in one secret or one actor, but in the way ordinary procedures could make troubling information disappear while allowing the organization to carry on.

How silence became part of the system

The investigation widened beyond individual allegations when reporters compared the ways sexual harassment claims were handled across different industries. Settlements and confidentiality clauses, already visible in reporting on Bill O'Reilly and Fox News, appeared not as rare exceptions but as common tools. Agreements could require women to surrender recordings, diaries, emails, and other evidence; limit their ability to assist others with similar claims; and require them to alert the accused and his representatives if a subpoena sought their testimony. Some provisions even barred a woman from responding to journalists or making disparaging statements about the host or network. In one agreement, a lawyer promised to advise O'Reilly on sexual-harassment matters, and a woman confirmed she had not filed a government complaint.

The sums involved show the bargain’s unequal stakes: one alleged victim received about $9 million and another $3.25 million, but either could lose the money by violating the restrictions. These agreements might compensate individuals, offer privacy, and let them move on. They could also strip away the evidence and testimony that would help other people recognize a repeated pattern. The claims and the material that might establish recurrence were placed beyond public view; the harasser could continue into a workplace where later employees lacked that history.

The book also explains why women and their lawyers accepted settlements. People might need the money, prefer privacy, see no better route, or want to avoid being branded liars, tattletales, flirts, or habitual litigators. Litigation offered no easy alternative: federal protections excluded some workers, including freelancers and employees of workplaces with fewer than fifteen people; the filing period could be as short as 180 days; and damages were capped at $300,000. Lawyers working on contingency generally received at least a third of an award and might receive nothing after a loss at trial. The system thus offered incentives to settle even when settlement terms suppressed information. The Equal Employment Opportunity Commission also kept many settlements confidential and could not publish employers’ records of complaints, leaving prospective employees unable to consult that information.

The Miramax settlement as a lived consequence

The experience of Zelda Perkins and Rowena Chiu makes the consequences concrete. After they resigned, they consulted a London attorney and initially expected criminal proceedings. Their lawyers told them they lacked physical evidence and had not contacted police in Venice, and recommended a settlement instead. Perkins and Chiu did not want personal payments; they proposed donating the money to charity as a way to make the matter public. They were told that Weinstein’s lawyers were unlikely to negotiate without a financial demand.

Perkins tried to make the agreement address future conduct as well as compensate them. Her proposals included therapy for Weinstein, a sexual-harassment policy and training at Miramax, and a process for evaluating complaints. After difficult negotiations, each woman received £125,000 in exchange for restrictions. Perkins later read from incomplete copies of the agreement: she could not discuss her time at Miramax, and even a medical professional she consulted about the events would have to sign a confidentiality agreement. The restrictions also required help in preventing further disclosure if information emerged. She had not been allowed to keep a complete copy; to consult the paperwork, she had to visit her lawyer’s office. Although Perkins and Chiu had separate but matching agreements, their confidentiality terms implied that they could not discuss the matter with one another.

The settlement’s effects extended beyond the document. Perkins could not readily explain to prospective employers why she had abruptly left a prominent company, and she concluded that her film career was over. She went to Guatemala to train horses. Her effort to secure therapy for Weinstein, including choosing a therapist, did not succeed. The agreement that was supposed to resolve a dispute instead narrowed the women’s ability to speak, impeded their professional lives, and failed to ensure the proposed changes occurred.

Oversight that protected the company

The same pattern appeared inside The Weinstein Company. Irwin Reiter, a longtime employee and the company’s most active board member, and Bob Weinstein became increasingly concerned about allegations during 2014 and 2015. Other employees knew little or nothing, but the company’s lawyers, assistants, contracts, and business practices could be turned toward furthering Weinstein’s conduct or hiding its consequences. Reiter had learned of an earlier assistant’s departure and settlement soon after joining Miramax in 1989; knowledge of a complaint did not by itself produce an effective intervention.

Lance Maerov had a formal watchdog role as one of three independent board members, appointed in 2013 to protect shareholders. At first he interpreted rumors about young women around Weinstein as evidence of infidelity and concentrated on financial misconduct and the company’s broader dysfunction. When an accusation of groping became public, he worried that a pattern could create liability and sought to examine Weinstein’s personnel file during contract renewal. Weinstein refused access, while David Boies offered to review the file himself and report possible legal problems to the board—a proposal Maerov distrusted because Boies’s work for the company and Weinstein appeared to overlap.

On July 1, 2015, Reiter and two executives gave Maerov a private look at material from the file over breakfast at the Four Seasons Hotel in Beverly Hills. Among the papers was a memo outlining what Weinstein had done to Emily Nestor. The executives feared Maerov would leave with the document, so he read it at the table. Yet seeing evidence did not bring a decisive response. Maerov later minimized the memo’s significance, and he did not press Boies for details after learning that women had received settlements.

A safeguard designed around liability

Four months later, Weinstein signed a new contract that secured his power for years. The contract introduced a code of conduct, allowed termination for misconduct, and required Weinstein to cover settlements arising from his misconduct, with escalating penalties: $250,000 for the first settlement, $500,000 for the second, and increasing amounts up to $1 million. These provisions acknowledged the possibility of further allegations, but they treated them as prospective financial costs to be allocated. Maerov’s primary concern was that the company not bear liability, not that women be protected from harassment. Once he believed the organization had legal and financial safeguards, he considered his intervention sufficient.

Reiter had tried to find a way to separate Weinstein from the company with Bob Weinstein, and had risked sharing information with a board member. Bob, meanwhile, accepted a private promise that Harvey would continue intensive therapy for sex addiction after Boies argued against putting the requirement in writing. A promise that could not be enforced took the place of the formal safeguards Bob had first wanted. The contract, the confidential assurance, and the board’s limited attention to evidence show different forms of failed oversight: information was available, but the measures adopted did not make protection of women the governing test.

Exposure and the limits of institutional self-protection

When The New York Times investigation was nearing publication, Maerov insisted on an emergency board call because Weinstein had hired lawyers to deal with a story the board did not know about. Boies told participants the forthcoming account would be damaging and described eight to twelve settlements, possibly more. He urged a united front rather than either defending or terminating Weinstein. The call brought board members into contact with the scale of the problem, but their discussion still turned heavily toward the company’s survival and responsibility. Bob Weinstein finally told his brother he was finished, while directors debated how to distance the company from past conduct and some hoped the scandal would be forgotten.

The pattern the book identifies is therefore broader than institutional ignorance: people could receive warnings, inspect evidence, and discuss allegations while responding in ways that preserved the institution or contained its exposure. The confidential settlements helped keep prior claims from informing later decisions; governance measures converted possible future misconduct into a schedule of costs; and formal promises displaced enforceable action. The investigation made those mechanisms visible as a connected system. It also shows why disclosure mattered: when claims and records were kept private, the public—and even people inside the relevant institutions—could not easily see how often the same problem had recurred.