For more than two decades, Alex Kozinski was one of the most powerful judges in America. As Chief Judge of the United States Court of Appeals for the Ninth Circuit — the largest and arguably most influential federal appellate court in the country, covering nine states and two territories and hearing more cases than any other circuit — Kozinski was a towering institutional presence: a prolific writer of opinions that shaped the law from free speech to intellectual property to criminal procedure, a celebrated lecturer, a mentor to hundreds of law clerks who regarded clerkships in his chambers as career-making opportunities, and a figure whose opinions on the law commanded attention at the highest levels of the legal profession.

He was also, according to dozens of women who worked for or alongside him, a serial sexual harasser who used the power of his position to subject female clerks, externs, lawyers, and court employees to decades of unwanted sexual conduct — conduct that ranged from displaying pornography on his chambers computer to touching women without their consent to making graphic sexual comments that left his targets shaken and professionally helpless.

In December 2017, after The Washington Post published a story in which fifteen women described Kozinski’s conduct in detail, he resigned from the federal bench. He was 67 years old. He had served for 35 years as a federal appellate judge. He resigned without a formal investigation, without a finding of misconduct, without criminal prosecution, and without bar discipline. He retired with his full federal pension intact — an annuity worth hundreds of thousands of dollars per year, funded by the taxpayers whose legal system he had corrupted through decades of abuse.

He has never been held criminally accountable. He has never been disbarred. He continues to write and lecture on the law.

The story of Alex Kozinski is not simply the story of one judge’s misconduct. It is the story of an institutional culture — the culture of the federal judiciary, with its lifetime tenure, its elaborate self-policing mechanisms, its close-knit hierarchy of clerks and judges, and its deep structural resistance to external accountability — that allowed his behavior to continue for decades because the costs of exposure fell entirely on the victims and the costs of silence fell on no one at all.

The Architecture of Power: Why Federal Clerks Cannot Fight Back

To understand why Alex Kozinski’s misconduct persisted for decades without effective institutional response, it is necessary to understand the structure of power that governs the relationship between federal appellate judges and their law clerks.

Federal appellate clerkships — particularly clerkships with judges on the Courts of Appeals — are among the most coveted positions in American law. They are obtained through a hyper-competitive selection process that begins in the third year of law school and involves recommendations from professors and law school deans, multi-round interviews, and selection decisions that are made entirely at the judge’s discretion. Clerking for a prominent circuit court judge — and particularly for a judge as prominent as Kozinski — opens doors to Supreme Court clerkships, elite law firm positions, prestigious academic placements, and careers at the highest levels of public law. A clerkship with Kozinski was widely regarded as a credential that could define the trajectory of a legal career.

The power differential this creates is almost impossible to overstate. A law clerk is typically 25 or 26 years old, at the beginning of a legal career, with enormous debt from law school and everything to gain from a successful clerkship. The judge is typically in his or her 50s or 60s, with decades of tenure, hundreds of former clerks spread across the legal landscape, and the ability to shape the young attorney’s prospects through recommendations, referrals, and the simple fact of institutional association. The clerk’s career depends, in a direct and not metaphorical sense, on the judge’s goodwill. The judge’s career depends on the clerk in no way at all.

When a judge uses this relationship to harass, the structural response available to the clerk is approximately zero. There is no human resources department. There is no EEO officer. The federal judiciary is not covered by Title VII of the Civil Rights Act — the primary federal statute that prohibits sexual harassment in employment — because Congress exempted itself and the judiciary from the law’s coverage when it passed the Civil Rights Act of 1964. The only formal recourse available to a harassed clerk is to file a complaint under the Judicial Conduct and Disability Act of 1980 — a law that routes complaints about a judge’s conduct to the chief judge of the circuit in which the judge sits, or, if the chief judge is the subject of the complaint, to the judicial council of the circuit.

Alex Kozinski was, for seven of the years during which his harassment was most extensively documented, the Chief Judge of the Ninth Circuit. He was the person to whom complaints about Ninth Circuit judges would be routed. He was responsible for the institution whose culture his own conduct was poisoning.

“The means of defence against foreign danger, have been always the instruments of tyranny at home. Among the Romans it was a standing maxim to excite a war, whenever a revolt was apprehended. Throughout all Europe, the armies kept up under the pretext of defending, have enslaved the people.”

What the Women Described: Years of Documented Misconduct

The Washington Post’s December 2017 report, by reporter Matt Zapotosky, was the product of months of reporting during which former clerks and court employees who had never spoken publicly about their experiences came forward to describe what had happened to them in Kozinski’s chambers. The fifteen women in the initial report described incidents that ranged across more than two decades and that were consistent in their pattern: Kozinski using the physical and social geography of his chambers to create situations in which he could subject women to unwanted sexual content or contact, often framed as humor or casual interaction, and relying on the power differential to ensure that the women felt they could neither object nor report.

Several women described being shown pornographic images on Kozinski’s chambers computer — images that he would display and then ask for their reactions, in a dynamic that left no good option: protest and risk career damage, or comply and participate in an interaction that felt degrading. Others described being groped — Kozinski allegedly putting his hands on their shoulders, backs, or legs in ways that were clearly not incidental. Several described incidents in which Kozinski made graphic sexual comments about their bodies or about sexual topics in ways that were calculated to shock and destabilize.

What multiple women emphasized in their accounts was not just the specific acts of harassment but the systemic nature of Kozinski’s approach. He was, multiple sources said, a man who tested boundaries deliberately and systematically, who understood that his position made retaliation essentially risk-free, and who had developed, over decades, a finely calibrated sense of how far he could push before encountering meaningful resistance. The harassment was not impulsive. It was practiced.

After the initial Post story, additional women came forward. By the time reporting was complete, more than 50 former clerks, externs, attorneys, and court employees had described experiences consistent with the pattern the original 15 had outlined. The incidents spanned the geographic range of the Ninth Circuit — Kozinski maintained chambers in Pasadena, California, and traveled to sittings at other Ninth Circuit locations — and the temporal range of his career. Women who had clerked for him in the late 1980s and women who had worked in the circuit in the 2010s described experiences that were, in their structural logic, essentially identical. Nothing had changed. No one had intervened. The culture had simply reproduced itself, year after year, administered by the same man in the same chambers, against the same category of vulnerable young lawyers.

The Whisper Network: How the Legal Community Knew and Did Nothing

Perhaps the most disturbing dimension of the Kozinski scandal — and the one with the most significant implications for institutional reform — is the evidence that his behavior was widely known within the legal community long before any public accounting occurred.

Former clerks describe a transmission of knowledge that passed from outgoing clerks to incoming ones as a kind of institutional orientation: when a new cohort of Kozinski clerks arrived, they were warned, informally and in private, about what to expect. Women clerks received specific warnings: be careful alone with him, do not look at his computer screen if he turns it toward you, know that certain kinds of comments will come and have a strategy for deflecting them. This knowledge was shared in confidence because sharing it publicly — or acting on it formally — would have ended careers. But it was shared. It was transmitted. It was a known and documented fact within the community of people who had experience with Kozinski’s chambers.

The knowledge extended beyond former clerks. Law professors at elite law schools who recommended students for Kozinski clerkships were aware of the whisper network. Some gave informal warnings alongside their recommendations. Others did not, concluding that the career benefits of the clerkship outweighed the risks they had heard about, or that the rumors were exaggerated, or that they did not want to be the person who cost a promising student a prestigious opportunity based on secondhand information. The law school pipeline — through which Kozinski recruited his clerks with the active assistance of professors at Yale, Harvard, Stanford, and other elite schools — continued to function even as knowledge of his conduct circulated at those schools.

The legal community knew. It processed that knowledge as a risk to be managed rather than a wrong to be corrected. It sent young women into his chambers with quiet warnings rather than public accountability. It allowed the institution to protect itself at the expense of every woman who walked through his door.

The 2008 Incident: A Pornography Website and a Self-Investigation

The Kozinski scandal had a precursor that, had it been handled differently, could have triggered the accountability reckoning that didn’t come until 2017. In 2008, the Los Angeles Times reported that Kozinski maintained a personal website — a file server accessible to family members — that contained sexually explicit and graphic material, including images of naked women and material that some viewers described as pornographic. Kozinski acknowledged the website but characterized the material as private humor shared with family, not pornography, and denied that it constituted judicial misconduct.

The complaint that followed was assigned for investigation to the Third Circuit Judicial Council — because the complaint could not be investigated by the Ninth Circuit, over which Kozinski then presided as Chief Judge. The Third Circuit investigation concluded that Kozinski had exercised “poor judgment” but had not committed conduct that warranted formal discipline. The complaint was closed.

The 2008 investigation was not conducted under circumstances that made a finding of serious misconduct likely. The materials on Kozinski’s website were reviewed in isolation from the pattern of conduct in his chambers that the 2017 reporting would expose. The investigation did not involve testimony from the women who had experienced harassment in his chambers — women who had never come forward because they had every reason to fear professional retaliation and no institutional mechanism to protect them. The self-policing framework of the federal judiciary, applied by a sister circuit whose judges had professional relationships and institutional incentives similar to those of the Ninth Circuit, produced a conclusion that protected the institution rather than the individuals the institution had harmed.

That 2008 investigation should have been a reckoning. It was a whitewash. And after it was closed, Kozinski continued for nine more years — continuing, by all accounts, to engage in the same conduct that had been known, whispered about, and tolerated since the beginning of his tenure.

Resignation Without Accountability: The Perfect Exit for the Powerful

On December 18, 2017, five days after the Washington Post published its initial report, Alex Kozinski announced his retirement from the federal bench, effective immediately. In a statement that demonstrated the particular quality of the non-apology available to the powerful, Kozinski said: “I have been a judge for 35 years and have always referred to my law clerks as my extended family. I would be devastated if I thought my sense of humor caused any of them serious harm.”

He expressed devastation at the possibility that his humor had caused harm. He did not acknowledge that he had harassed anyone. He did not apologize to any specific person. He did not acknowledge that the dozens of women who had described their experiences with him were telling the truth. He retired, instead of being removed, and the retirement allowed him to collect his full federal judicial pension — a benefit calculated on the basis of his years of service and his judicial salary, which had been approximately $240,000 per year at the time of his resignation.

Under the federal judicial retirement statute, 28 U.S.C. § 371, a federal judge who retires after qualifying for senior status receives retirement pay equal to the salary they were receiving at the time of retirement. Kozinski, having served for more than 15 years and having reached the age threshold required for full retirement benefits, was entitled to retire with full pay. His resignation, rather than removal, preserved that entitlement. Had he been removed through impeachment proceedings, he would have lost his pension. By resigning before any formal proceeding could be completed, he secured his financial future at taxpayer expense.

The Judicial Council Reform and Judicial Conduct and Disability Act of 1980 provides that when a judge who is the subject of a judicial conduct complaint retires or resigns, the complaint proceedings terminate. The law was designed to address the problem of a judge who retires to avoid discipline for minor misconduct. Applied to Kozinski’s situation, it operated as a mechanism for extinguishing the possibility of any formal finding about his conduct: once he resigned, the pending investigation was closed without a conclusion, without findings, and without any public accounting of what the investigation had found.

“If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.”

The Title VII Gap: How Congress Exempted the Judiciary from Sexual Harassment Law

The structural gap that most directly enabled Kozinski’s conduct to persist without legal consequence is also the most politically embarrassing: the federal judiciary is not subject to Title VII of the Civil Rights Act of 1964, the primary federal statute that makes sexual harassment in employment illegal.

Title VII prohibits discrimination on the basis of sex, race, color, national origin, and religion by “employers” as defined by the statute. The statute defines “employer” to exclude the United States government and its subdivisions, including the federal judiciary. Federal employees who are harassed by their supervisors must instead rely on Title VII as applied to federal agencies through a separate provision — Section 717 of the Civil Rights Act — which requires exhaustion of administrative remedies through the Equal Employment Opportunity Commission before a lawsuit can be filed. But Section 717 has been interpreted not to apply to law clerks and other judicial employees, because the federal judiciary is treated as a unit of the judicial branch separate from the executive-branch agencies for which the EEOC process was designed.

The practical result is that a law clerk who is sexually harassed by a federal judge has no civil legal remedy under Title VII, no EEOC complaint mechanism, and no federal court lawsuit available to them — because federal courts have consistently declined to exercise jurisdiction over claims by judicial employees against sitting federal judges for conduct occurring in the judicial employment relationship. The irony of this situation is not subtle: the courts that would have to adjudicate a harassment claim against a federal judge are staffed by federal judges, in a system with powerful institutional incentives to protect its own members from civil liability.

This gap was known to Congress for decades. The Congressional Accountability Act of 1995 extended civil rights protections to Congressional employees, acknowledging that the legislature had previously exempted itself from employment discrimination law. No equivalent legislation extended similar protections to judicial employees. The judiciary’s independence was treated as a reason to exempt it from the laws that govern every other workplace in America.

In the aftermath of the Kozinski scandal and the broader #MeToo reckoning of 2017-2018, the federal judiciary announced several internal reforms: the creation of a confidential employee dispute resolution program, clearer guidance on what constitutes misconduct, and an expansion of the formal complaint process. The Judicial Conference issued updated model workplace conduct policies. Some circuits hired employee relations officers. These reforms addressed the surface-level procedural gap without touching the fundamental structural problem: judicial employees still lack the civil legal remedies available to every other class of American worker, and the self-policing framework that failed for decades remained, in its essential architecture, intact.

The Clerks Who Could Not Speak: Career Destruction as Silence Mechanism

The women who experienced Kozinski’s harassment and chose silence — the majority, over decades — were not making a simple or a cowardly choice. They were making a rational calculation about professional survival in a system that left them no good options.

Speaking publicly about a federal appellate judge’s misconduct, for a young attorney at the beginning of their career, was not merely risky. It was potentially career-ending. The federal judiciary is a small world. Federal appellate judges know each other. They communicate about clerks — formally, through recommendation letters, and informally, through the social networks that develop across decades of shared institutional life. A clerk who accused a sitting federal appellate judge of sexual harassment would face, in addition to whatever formal or informal retaliation the accused judge could orchestrate, the simple fact of being known, throughout the federal judicial community, as the person who had made such an accusation. In a world where career advancement depended on the goodwill of federal judges — for clerkship recommendations, for appointments to positions in the executive branch, for nominations to federal judicial positions, for referrals of major cases — that identification was a professional death sentence.

Several of the women who ultimately spoke to the Washington Post in 2017 had achieved sufficient career security that the professional cost of speaking had declined to a level they were willing to bear. They were established attorneys, law professors, partners at major firms. They had built careers that did not depend on Kozinski’s goodwill. The women who had come through his chambers most recently — the ones most exposed to ongoing harm, the ones for whom the whisper network was most immediately relevant — remained, for the most part, silent. Their silence was rational. It was also a measure of the extent to which the institutional system had structured its incentives to protect the judge rather than the people he harassed.

Kozinski’s Legal Legacy: Can You Separate the Judge from the Work?

One of the most vexed questions that the Kozinski scandal raised — and has not fully answered — is what to do with the legal legacy of a judge who was also a systematic harasser. Kozinski authored hundreds of significant opinions during his 35 years on the Ninth Circuit. He wrote influential decisions in copyright law, free speech doctrine, criminal procedure, and administrative law. His opinions are cited in law school casebooks, in treatises, in the briefs of practitioners across the country. His former clerks — including those who described his harassment — are judges, professors, and senior lawyers whose professional formation he shaped.

The question of whether the intellectual legacy of a powerful abuser can be cleanly separated from the conduct that the power enabled does not have a satisfying answer. Kozinski’s opinions do not become legally incorrect because of his personal misconduct. The First Amendment doctrine he helped shape is not invalidated by his behavior in chambers. But the erasure of the connection — the treatment of his judicial output as if it existed in a separate universe from his conduct — is itself a kind of institutional protection of the powerful. The clerks who did some of the work reflected in those opinions were working in an environment of harassment and fear. The intellectual productivity that made Kozinski valuable to the institution was produced, at least in part, by people who were simultaneously being harmed by him.

This question is not merely abstract. Former Kozinski clerks who were victims of his harassment have had to navigate a legal world in which his work is pervasive — in which the cases they argue cite his opinions, in which their own professional training was shaped by a man who abused his relationship with them, in which the institutional prestige of the clerkship that appears on their resume is inseparable from the experience of the harassment. The legacy question is not academic for them. It is a daily confrontation with the way institutional power launders individual harm.

The Broader Federal Judiciary #MeToo Reckoning: One Judge Is Not the Story

The Kozinski scandal was the most prominent disclosure of a broader pattern that the #MeToo moment made temporarily visible in the federal judiciary. In the months following Kozinski’s resignation, other federal judges faced public accusations of harassment or misconduct, and the conversation about the structural vulnerabilities of the federal judicial employment relationship opened briefly before closing again under the weight of institutional inertia.

A 2018 survey of more than 800 current and former federal law clerks, conducted by researchers at Harvard Law School and reported in a law review article, found that approximately one in four female respondents reported experiencing some form of sexual harassment during their clerkship. One in six reported behavior that met the legal definition of a hostile work environment. The survey was not exhaustive, and its findings understate the problem to the extent that current clerks — who remain in the most professionally vulnerable positions — were underrepresented among respondents. But the order of magnitude is clear: Kozinski was not an anomaly. He was a prominent case in a distribution of misconduct whose full extent the judiciary’s self-policing structure had systematically concealed.

Federal District Judge Stephen Reinhardt — like Kozinski a towering figure of the Ninth Circuit whose death in 2018 preceded any formal accountability for allegations of workplace misconduct that had circulated for years — illustrated the broader pattern. Multiple clerks and employees described a chambers culture at Reinhardt’s chambers that was, in different ways, characterized by the same structural problems: the power differential, the career dependency, and the absence of any safe mechanism for raising concerns.

Federal District Judge Carlos Murguía of the District of Kansas resigned in 2018 after BuzzFeed News reported that he had engaged in an ongoing affair with a woman who appeared before his court while her ex-husband’s probation violation was pending. The pattern — a judge using the power and prestige of the office to exploit people whose professional or legal lives fell within his jurisdiction — was structurally identical to the Kozinski pattern, even as the specific form of conduct differed.

The federal judiciary’s response to these revelations was, characteristically, internal, procedural, and protective of the institution’s autonomy. Chief Justice John Roberts, in his 2017 Year-End Report on the Federal Judiciary, acknowledged that the judiciary was “not immune from the problem” of workplace harassment and announced a review of existing procedures. The resulting reforms — updated workplace conduct policies, expanded dispute resolution mechanisms, enhanced training requirements — were real improvements at the margin. They did not address the fundamental structural problems: the absence of civil legal remedies for judicial employees, the self-policing framework that routes complaints back through the same institutional hierarchy that created the conditions for misconduct, and the lifetime tenure that insulates federal judges from the accountability mechanisms available for virtually every other category of public employee.

The Pension Question: Public Funds for Private Abuse

The question of Kozinski’s pension is not a minor detail. It is the most concrete expression of the accountability gap at the heart of federal judicial misconduct.

Kozinski resigned in December 2017 and immediately became eligible for retirement pay under 28 U.S.C. § 371(c), which allows federal judges to retire with full pay equal to their active judicial salary after reaching age 65 with at least 15 years of service. Kozinski met both thresholds. His retirement pay, calculated on the basis of a Circuit Judge salary of approximately $240,000 per year, is a lifetime annuity that will cost taxpayers hundreds of thousands of dollars annually for as long as he lives. This annuity is funded by the same taxpayers whose daughters and nieces and students clerked in his chambers and were harassed there.

Federal law provides no mechanism for forfeiting or reducing a retired federal judge’s pension as a consequence of misconduct discovered after retirement, unless the judge has been criminally convicted of an offense specifically listed in the forfeiture statute. Sexual harassment, however severe and however extensively documented, is not an offense that triggers pension forfeiture absent a criminal conviction. The Judicial Council Reform Act’s provision terminating misconduct proceedings upon a judge’s retirement means that even a formal judicial misconduct finding — which was never made in Kozinski’s case — would have had no effect on his pension.

The result is a system in which a federal judge can harass dozens of women over decades, resign to avoid formal accountability, and retire on a full government pension — and there is no legal mechanism, short of congressional action, to address any of these outcomes. This is not a gap created by oversight. It is the predictable result of a system designed, from the ground up, to protect the institution and its members from external accountability.

Reform Blueprint: Making Federal Judicial Accountability Real

1. Extend Title VII and the Civil Rights Act to cover all federal judicial employees, including law clerks, externs, and court staff. Congress should pass legislation explicitly extending Title VII’s employment discrimination and sexual harassment protections to judicial branch employees, creating a private right of action enforceable in federal court. The legislation should include independent adjudication of claims involving federal judges — before a panel drawn from circuits other than the one in which the judge sits, with EEOC oversight of the filing and processing of complaints — to avoid the conflict of interest inherent in allowing judges to adjudicate claims against their colleagues. The judiciary’s independence does not require exemption from the same employment law that governs every other workplace in America.
2. Reform the Judicial Conduct and Disability Act to allow misconduct proceedings to continue after a judge’s resignation or retirement. The current provision automatically terminating misconduct proceedings upon a judge’s retirement should be repealed and replaced with a provision permitting continuation of proceedings at the discretion of the investigating judicial council, with a finding that the misconduct was of sufficient severity to warrant public disclosure even in the absence of a sitting judge. Completed findings of serious misconduct should be published and made part of the permanent public record of the federal judiciary, regardless of whether the subject judge remains on the bench.
3. Create pension forfeiture provisions for federal judges found to have engaged in serious sexual harassment or misconduct. Congress should amend the federal judicial retirement statute to provide for forfeiture or reduction of retirement benefits in cases where a judicial conduct proceeding — whether completed before or after retirement — results in a finding of serious misconduct. At minimum, judges who resign while under investigation for conduct that would have warranted removal should not be entitled to the same retirement benefits as judges who serve with distinction. The pension system should not function as a financial reward for successful evasion of accountability.
4. Establish an independent Office of Judicial Integrity within the judicial branch, with external oversight from both Congress and the executive branch. The self-policing structure of the federal judiciary — in which complaints about a judge are investigated by that judge’s colleagues in the same circuit — is structurally incapable of providing the kind of independent accountability that serious misconduct demands. An independent Office of Judicial Integrity, staffed by non-judge professionals with backgrounds in employment law and civil rights investigation, should be established within the judicial branch but subject to oversight by a joint congressional committee and with mandatory reporting requirements to the Department of Justice for conduct that may constitute criminal violations. Investigators in this office should have subpoena authority to compel testimony and document production from judicial chambers.
5. Mandate anonymous reporting mechanisms and create formal whistleblower protections for judicial employees who report misconduct. Every federal circuit should be required to maintain an anonymous, electronically accessible reporting mechanism for judicial employees who wish to report misconduct without identifying themselves. Reports to this mechanism should be reviewed by the Office of Judicial Integrity, not by the circuit’s chief judge. Judicial employees who report misconduct through formal channels should be entitled to the same whistleblower protections available to other federal employees under the Whistleblower Protection Act, with private rights of action for retaliation and with independent adjudication of retaliation claims by panels outside the circuit in which the alleged retaliation occurred.
6. Require transparency in the handling of all judicial conduct complaints, with public reporting of outcomes. The current system processes most judicial conduct complaints in secrecy, with neither the complainant nor the public informed of the outcome of proceedings. Congress should amend the Judicial Conduct and Disability Act to require public reporting of all complaints that are not dismissed at the initial review stage, with anonymized summaries of findings in cases that are closed without formal action and full public disclosure of findings in cases resulting in formal discipline. The secrecy of the current process is not justified by any legitimate interest in judicial independence; it is justified only by the institutional interest in protecting judges from accountability.
7. Reform the law clerk selection system to reduce the power imbalance and create formal evaluation mechanisms for chambers culture. The current system, in which law clerks are selected by individual judges through an unregulated, highly informal process, maximizes the power differential that enables harassment. Congress and the Judicial Conference should work toward a more structured selection process with oversight, formal evaluation surveys of clerks at the conclusion of their clerkships, and aggregated analysis of responses to identify chambers with patterns of concerning conduct. Law schools that participate in the clerkship placement process should require candidates to acknowledge that reporting mechanisms are available and should maintain formal channels for receiving reports from students who experience misconduct during clerkships.

The Institution That Protected Itself

Alex Kozinski retired in December 2017 with his pension, his legal legacy, and his freedom intact. The women who had been harassed in his chambers over more than three decades retired with something less: careers shadowed by an experience they had been required to keep silent, professional identities shaped partly by the months they spent in a chambers where the price of admission was tolerance of unwanted sexual conduct, and the knowledge that they had served in an institution that valued the man who harmed them more than it valued them.

The federal judiciary did not cover up Kozinski’s misconduct in the sense of actively concealing specific incidents. It covered it up in a more fundamental and more insidious sense: it built and maintained a system in which the incentives for silence vastly outweighed the incentives for disclosure, in which the self-policing mechanisms were structurally incapable of acting on conduct patterns that were whispered about for decades, and in which the institutional response to exposure was resignation — with full benefits — rather than accountability.

The Founders who designed the federal judiciary’s structure of lifetime tenure and salary protection intended those protections to insulate judges from political pressure — to ensure that judges could rule against popular majorities, against powerful interests, and against the government itself without fear of losing their positions. They did not design those protections to insulate judges who harassed their employees. They did not contemplate that the mechanism created to protect judicial independence would be weaponized to protect judicial misconduct.

The reforms that would fix what Kozinski’s tenure exposed are not complicated. They require only the political will to acknowledge that the federal judiciary is an institution staffed by human beings who are capable of the same abuses of power as any other class of official — and that the accountability mechanisms designed for those officials must apply to judges as well. Until those reforms are made, the whisper networks will continue to transmit their warnings. The clerks who receive those warnings will continue to make their impossible calculations about silence and career. And the men who created those networks will continue to retire with their pensions and their legacies, while the institution looks the other way.