The most power, the least accountability
Judges and prosecutors can jail you, take your children, and end your life through the state. When they abuse that power, the system that’s supposed to check them disciplines almost no one — and the law itself shields the worst of them from being sued.
This hub documents the accountability gap from both sides of the courtroom: the steep funnel between misconduct complaints and actual discipline, the scale of prosecutorial misconduct in the cases we know were wrong, and the doctrine of absolute immunity that lets a prosecutor who hides evidence or uses perjury walk away untouched. Every figure is graded fact and pinned to the courts’ own data, landmark reporting, and Supreme Court law — and every number is given a face.
What this hub is about
The judiciary and the prosecution are the two offices with the most direct power over an individual’s freedom — and the two that are, in practice, the least accountable when they misuse it. Thousands of misconduct complaints are filed against judges every year and almost none end in public discipline; prosecutorial misconduct shows up in a majority of the convictions later proven wrong; and a Supreme Court doctrine of absolute immunity means a prosecutor who deliberately hides evidence or puts on perjured testimony cannot be sued for it.
We build this the way we build everything: the aggregate numbers, graded and pinned to primary sources, and then the specific human cases that show what the numbers cost. The point is not that most judges or prosecutors are corrupt — the large majority aren’t. It’s that the mechanisms meant to catch and punish the ones who are barely function, and are shielded by secrecy and immunity by design.
The hub launches with three graded spokes:
The misconduct funnel. More than 1,850 federal complaints in a year (86% challenging a ruling, not conduct); a state commission that took 675 complaints and produced a single public sanction; Reuters’ finding of at least 5,206 people harmed and 9 in 10 disciplined judges kept on the bench. Read the numbers.
Prosecutorial misconduct. 550+ capital cases reversed for prosecutorial misconduct (DPIC); official misconduct in 54% of all exonerations (National Registry); Michael Morton’s 25 years while the evidence sat in the file, and Curtis Flowers tried six times by the same prosecutor — who paid nothing. Read the record.
Absolute immunity. Since Imbler v. Pachtman (1976), prosecutors cannot be sued for trial conduct even when they knowingly break the law — the doctrine that let the office which buried evidence and sent John Thompson to death row for 14 years escape a $14 million judgment. Read the doctrine.
The human cases. “Kids for Cash” (two PA judges, ~$2.6M in kickbacks, thousands of children jailed); Judge Les Hayes’s 496-day sentence for unpaid traffic tickets; Morton; Flowers; Thompson. The faces behind the funnel.
OPENSpecific, documented cases from readers — a named judge or prosecutor with a paper trail (an opinion, a bar record, a lawsuit) — to build into individual spokes.
OPENThe confidential majority: what’s inside the private-sanction systems of the 38–44 states that keep judicial discipline secret.
OPENA true national count of prosecutors actually disciplined versus the number of court-found misconduct cases — no comprehensive registry exists.
OPENElected judges and campaign money — how judicial elections and donor relationships bear on the appearance of impartiality.
This is a living record, and our readers make it stronger. If there is a thread here you want us to chase, or you have documents or firsthand knowledge that would sharpen a claim, send it to us. We read every lead.
The same hub, restaged one beat at a time. Step through it here, or present it fullscreen.
Judges & Prosecutors.
The most power over your liberty, the least accountability when they abuse it — the discipline funnel, prosecutorial misconduct at scale, and the immunity that shields it.
How accountability is engineered out
The funnel. Complaints pour in and almost nothing comes out the other end. Much of that is legitimate — disciplinary bodies review conduct, not whether a ruling was wrong — but the cases that do clear that bar are handled largely in secret, by other judges. Public reprimand or removal is a rounding error.
The secrecy. At least 38 states (44 by the National Center for State Courts’ current count) issue private sanctions in which the judge’s name and offense never become public. “Judges judging judges,” as one ethics scholar puts it, tends toward leniency.
The immunity. For prosecutors, the shield is written into constitutional law. Absolute immunity means that even deliberately hiding exculpatory evidence or using perjured testimony carries no civil liability. The ordinary consequence of proven, court-confirmed misconduct is, at most, a reversed conviction — a remedy for the victim, not a sanction for the official.
The cost. It lands on real people: children funneled into private jails, a mother imprisoned 496 days over traffic tickets, Michael Morton’s 25 years, John Thompson’s 14 on death row. When accountability is this rare, the harm is effectively licensed.
The numbers, then the faces — and the fair objections
Every figure is graded FACT and pinned to a primary source: the Administrative Office of the U.S. Courts, state judicial-conduct commissions, the Death Penalty Information Center, the National Registry of Exonerations, Reuters’ “Teflon Robe,” and Supreme Court opinions. We steelman the system where it deserves it — most complaint dismissals are proper, and most judges and prosecutors are not corrupt — and we refuse to invent a single national rate the data can’t support. Where the defensible facts run out but the question is serious, we pose it as a question worth demanding an answer to, rather than asserting past the evidence.
The record, piece by piece
The hub launches with its numbers-and-doctrine backbone; the open leads above are the next spokes, each built to the same graded, sourced standard.
The misconduct funnel: how judicial complaints vanish
A high-level, by-the-numbers look at how rarely judicial misconduct is publicly punished — the funnel between complaints filed and judges disciplined. Federal: the Administrative Office of the U.S. Courts logged 1,850+ misconduct complaints against federal judges in the 12-month period ending September 30, 2025 (a ~23% increase); litigants filed 79%, and 86% of allegations challenged the merits of a ruling (dismissed as a matter of course, since disciplinary panels review conduct, not legal error). State: North Carolina's Judicial Standards Commission received 675 new complaints in 2024 (up from 653 in 2023 and 470 in 2022); of 753 pending, 633 were dismissed after initial review and 32 more after preliminary investigation; 4 statements of charges and 3 public-discipline recommendations issued, and the NC Supreme Court handed down a single disciplinary decision (a 120-day suspension). States handle ~98% of U.S. litigation before ~30,000 state judges (NCSC). Reuters' 2020 investigation 'The Teflon Robe' found at least 5,206 people directly harmed by judges' misconduct over ~12 years, at least 341 judges who escaped discipline by resigning/retiring amid allegations, and that at least 38 states issue secret/private sanctions with judges largely policed by other judges ('judges judging judges,' per NYU's Stephen Gillers). We grade every figure FACT and attribute it to the courts' own statistics or Reuters; we steelman the legitimate merits-exclusion rule (most dismissals are proper) rather than imply every dismissal is a cover-up, and we assert no single national rate because the systems aren't aggregated and much of the record is confidential. Home is provisional (The Untouchables); the piece is cross-linkable to Policing the Poor and the Israeli Influence judge case, and prosecutorial misconduct is the planned next spoke.
Prosecutorial misconduct: documented at scale, punished almost never
The prosecution-side companion to the judicial-misconduct funnel. The scale of prosecutorial misconduct is documented in the cases proven wrong: the Death Penalty Information Center identified more than 550 capital cases reversed for prosecutorial misconduct or ending in a misconduct exoneration since 1972 (>5.6% of all death sentences, across 228 counties, 32 states, and federal prosecutions), with withholding favorable evidence (35%) and improper argument (33%) the most common forms, and 69% of death-row exonerations involving official misconduct. The National Registry of Exonerations' 2020 report 'Government Misconduct and Convicting the Innocent' found official misconduct by police or prosecutors in 54% of exonerations (higher for murder cases). The anchor case: Michael Morton, convicted in 1987 of murdering his wife and imprisoned nearly 25 years while prosecutor Ken Anderson withheld exculpatory evidence (the couple's 3-year-old son's account that 'a monster,' not his father, was the killer; a green van behind the house; the stolen purse and fraudulently used credit card; unidentified prints). DNA freed Morton in 2011 and identified the real killer, Mark Alan Norwood. A Texas court of inquiry found Anderson had concealed evidence; in 2013 he pleaded to criminal contempt, surrendered his law license, and was jailed with 500 hours of community service — widely called the first U.S. prosecutor jailed for such misconduct — and Texas passed the Michael Morton Act mandating open-file discovery. The impunity is structural: since Imbler v. Pachtman (1976) prosecutors have absolute immunity from civil suit for trial conduct, and state-bar discipline is documented to be extraordinarily rare. We grade every figure FACT and attribute the counts to DPIC and the National Registry; we assert no single national misconduct or discipline rate and do NOT claim most prosecutors offend — the defensible point is that when misconduct is proven, a personal consequence for the prosecutor is the rare exception. Home is provisional (The Untouchables); cross-linkable to Policing the Poor and paired with the judicial-misconduct funnel.
Absolute immunity: the shield for prosecutors who break the law
Absolute prosecutorial immunity is a judge-made doctrine barring civil suits for money damages against prosecutors for conduct tied to the 'judicial phase' of a case — charging, presenting evidence, questioning witnesses — even when it is intentional and unlawful. Established in Imbler v. Pachtman (1976), it means a prosecutor who knowingly uses perjured testimony or hides exculpatory evidence at trial cannot be sued for it; only investigative or public-relations acts (advising a raid, a press conference, pre-warrant work) get the weaker qualified immunity (Burns v. Reed, 1991; Buckley v. Fitzsimmons, 1993). The starkest illustration is Connick v. Thompson (2011): John Thompson spent 18 years in prison, 14 on death row, after Orleans Parish prosecutors suppressed a blood report that cleared him; a jury awarded $14 million and the Supreme Court reversed it 5-4 (Thomas majority; Ginsburg dissent), holding an office can't be liable under Section 1983 for a single Brady violation absent a proven pattern. The Court defends the doctrine (preventing retaliatory suits; avoiding a chilling effect; pointing to criminal charges, elections, and bar discipline as alternatives) — but the alternatives barely operate: the Center for Prosecutor Integrity's 2013 report found that of 3,625 documented misconduct cases, fewer than 2% of prosecutors faced any public discipline and virtually none faced criminal charges. We grade the doctrine, cases, rationale, and the CPI figure as FACT, and pose the objections (why shield knowing fabrication; why not a narrow process to prove knowing misconduct on new evidence; why rely on remedies that don't operate) as open questions — not legal conclusions. We do NOT claim immunity has no legitimate purpose or that most prosecutors abuse it.
The taxpayer bill: who actually pays for misconduct
The public-cost dimension of the accountability gap. When official misconduct violates constitutional rights, compensation runs through a civil-rights suit under 42 U.S.C. § 1983 — but the individuals are shielded (police by qualified immunity, prosecutors by absolute immunity), so suits target the municipality, which under Monell v. Department of Social Services (1978) is liable only for a policy, custom, or failure to train. The result: taxpayers, not wrongdoers, pay. Chicago's Department of Law analysis (a consent-decree requirement) found the city spent nearly $259 million in 2025 resolving police-misconduct lawsuits — more than 3x the 2024 total — with wrongful-conviction cases accounting for $193.4 million (~75%) and vehicle pursuits $54.4 million. The Legal Aid Society's analysis of city data found New York City paid $117.25 million in 2025 to settle NYPD-misconduct claims and more than $796 million since 2019 (a floor, since it excludes pre-litigation Comptroller settlements). These municipal totals are primarily POLICE misconduct; the wrongful-conviction share is where prosecutorial misconduct (fabricated/withheld evidence) enters, usually bundled with the police work — we say so and do not break out a prosecutor-only total the public data can't support. And the bill understates the harm: absolute immunity plus Connick v. Thompson (2011) means that in the hardest prosecutorial cases neither the official nor the office nor the treasury pays, and the loss stays with the victim (John Thompson recovered nothing after 14 years on death row). We grade the figures and the § 1983/Monell mechanism as FACT; we assert no national total (none exists) and do not claim every settlement reflects proven misconduct.
“We close at 5”: Judge Sharon Keller and the discipline that wasn’t
A named-case spoke: the discipline funnel made flesh. On September 25, 2007 — the day the U.S. Supreme Court agreed to hear a challenge to lethal injection (Baze v. Rees) — lawyers for Texas death-row inmate Michael Richard suffered computer failures and asked the Texas Court of Criminal Appeals to keep its clerk's office open briefly past 5 p.m. to accept a filing; presiding judge Sharon Keller refused ('We close at 5'), and Richard was executed that night without the court considering the claim. The refusal drew cross-ideological condemnation. The Texas State Commission on Judicial Conduct charged Keller and issued a public warning in 2009; a special court of review threw it out in 2010 on procedural grounds, leaving her neither formally sanctioned nor exonerated, and she remained presiding judge of the state's highest criminal court through 2024. Separately, the Texas Ethics Commission fined her about $100,000 in 2010 for failing to disclose roughly $2 million in real estate. It was not out of character: in a 1996 opinion (Cesar Fierro) she acknowledged a due-process violation from a confession likely coerced by threats to torture the defendant's parents — with a detective later admitting perjury — yet held the error 'harmless' and denied a new trial the prosecutor and trial judge agreed he deserved. We grade the documented facts FACT, report the (non-)consequence, and do NOT assert a surviving misconduct finding (the warning was vacated) or that any ruling was corrupt.
Elected judges and the money that funds them
The United States is nearly alone among major democracies in electing judges: roughly 38 states choose or retain at least some of their judges by election, which forces judges to raise campaign money — often from the lawyers, litigants, and industries who appear before them — straining the appearance of impartiality that defines a court. The clearest illustration is Caperton v. Massey (2009). After a West Virginia jury awarded Hugh Caperton's coal company $50 million for Massey Coal's fraud (2002), Massey CEO Don Blankenship spent more than $3 million — about 3,000 times the legal maximum for a direct contribution, and more than all of Brent Benjamin's other supporters and his own campaign combined — to help elect Benjamin to the West Virginia Supreme Court of Appeals; Benjamin then refused Caperton's recusal request and joined the 3-2 majority overturning the $50 million verdict. The U.S. Supreme Court held 5-4 that the Due Process Clause required Benjamin's recusal because such 'extreme facts' create a 'probability of bias.' But Caperton drew the constitutional line only at the extreme; below it, judges in most states still rule on their own recusal motions and money keeps pouring into judicial races (a trend the Brennan Center tracks). We grade the doctrine and the Caperton facts as FACT, attribute the scale-of-spending trend to the Brennan Center rather than invent a total, and do NOT claim elected judges are generally corrupt or that ordinary donations fix cases below the Caperton threshold — the issue is the structural appearance problem and where the Court drew the line.
Disciplined in the dark: the secret-sanction states
The second, quieter failure of the judicial-misconduct funnel: what happens to the cases that survive review. By the National Center for State Courts' current count, 44 states issue private dispositions or sanctions for judicial misconduct — private admonitions, letters of caution, confidential agreements — in which the judge's identity and the nature of the offense are never made public; the point at which confidentiality lifts (if ever) varies enormously by state, and some publish only aggregate counts or almost nothing. The scale is measurable: Reuters' comprehensive investigation identified 3,613 cases from 2008-2018 in which states disciplined judges but hid key details, including in many the judges' identities, from the public — on top of the 1,509 publicly disciplined cases it reviewed, of which 9 in 10 judges returned to the bench. Judicial-ethics scholars (e.g., NYU's Stephen Gillers) note that private discipline destroys the deterrent effect, because other judges never learn what conduct was punished. We grade the NCSC and Reuters figures FACT, steelman the legitimate purpose of confidentiality (protecting judges from frivolous or unproven complaints), and assert no single national figure the fragmented systems can't support — the point is that on the proven end the public is structurally prevented from seeing how judicial misconduct is handled. The NCSC's interactive database is the best public tool for comparing what each state discloses and when.
Sibley Reynolds
A Judges & Prosecutors case graded to a single documented act, with the louder online material deliberately excluded. FACT: the Montgomery Advertiser reported in 2016 that Chilton County, Alabama Circuit Judge Sibley Reynolds held Carlos Ortigoza — an Air Force National Guard veteran — in contempt of court and sent him to the county jail after Ortigoza violated the judge's order not to post on social media about his own divorce case by publishing a post criticizing the judge; the target of the criticism, the author of the order, and the official who ordered the jailing were the same person. PROBABLY TRUE (posed, not decided): that the gag order and the jailing were an unconstitutional prior restraint — criticizing a judge is long-settled protected speech and orders forbidding a party from speaking publicly draw the heaviest scrutiny and are presumptively invalid, but the specific order was never appealed, so no court ruled on it, which is itself part of the story. FACT (structural): what makes such an act nearly uncheckable is the system the judge operates inside — the contempt power lets a judge act as accuser, witness, and judge of an offense against the court's own authority and jail for it with little process; absolute judicial immunity bars suing a judge for a judicial act even if unlawful, leaving a wrongly jailed citizen no civil remedy against him; and in many states judicial discipline is confidential, so the public and future voters never learn of it. Fenced out entirely: the broader viral allegations about this judge — custody rulings, 'child abuse,' bias, property — which circulate on parents'-rights Facebook pages and crowd-sourced registries that state plainly their entries are 'reports from families, not court findings'; we do not assert any of them. The hardest question the case forces: why can a single official be the target of your speech, the author of the order banning it, and the one who jails you for it, all at once — and, shielded by immunity and secret discipline, answer to no one for it?
A Crime to Repeat It
A press-freedom / prior-restraint story, graded so the point survives any fight over the underlying scandal. FACT: in October 2026 the Republican Party of Florida filed a verified complaint in Highlands County circuit court against the Florida Democratic Party, its candidates for governor (David Jolly), attorney general (Jose Javier Rodriguez), and chief financial officer (Annette Taddeo), and the owners of 11 Orlando and Tampa Bay TV stations (including WFTV, WESH, WKMG, WOFL, WTVT, WFLA, WTSP, WFTS), seeking to bar a 30-second ad before the defendants could respond — and referred the stations to FDLE and prosecutors (Tampa Bay Times; WUSF/Florida Trident; Miami Herald). FACT: the legal theory is that, because a state appeals court ordered the grand jury report expunged on Oct. 7, broadcasting its 'content, gist, or import' is now a first-degree misdemeanor and criminal contempt under Fla. Stat. 905.27, punishable by up to a year in jail; party lawyer Benjamin Gibson wrote to station managers that 'knowingly broadcasting advertisements built on the contents of the confidential Presentment is criminal conduct in Florida,' and the injunction sought would bar 'any other advertisement or material' relaying the report's gist. FACT: the ad repeats the statewide grand jury's own finding that $10 million from the state's Centene Medicaid settlement was 'misappropriated' and routed into a political committee tied to Attorney General James Uthmeier (now on the Nov. 3 ballot). FACT (attributed to the grand jury, NOT graded as an adjudicated crime): the misappropriation finding is a grand jury presentment, not a conviction; DeSantis and Uthmeier deny wrongdoing; the report was expunged — carried. FACT: the party sued before the expungement order was even final (a 15-day window meant it would not take effect until roughly Oct. 23), about a report that had been public since August; RPOF Chair Evan Power, asked if suing was premature, said 'they're not going to change their minds.' FACT: the criminal referral was pointed in part at statewide prosecutor Brad McVay, who works inside Uthmeier's own office — the target's office asked to weigh charges over an ad naming the target. PROBABLY TRUE (attributed to free-speech experts and media attorneys, since no court has finally ruled): the move is a textbook prior restraint facing steep First Amendment hurdles — and a Highlands County judge already denied the emergency no-notice request. The thesis: whatever the truth of the scandal, a political party is trying to make it a crime to repeat a grand jury's finding — on a non-final order, about an already-public report, with the referral aimed at the accused's own office — a template that, if it worked, would let any embarrassed administration expunge a report and then criminalize its citizens' memory of it. The hardest question: if a true finding can be erased and then made illegal to repeat, what is left of the difference between correcting the record and covering it up?
One Hundred Percent Certain
The graded record of Dr. James Grigson, the Dallas forensic psychiatrist nicknamed 'Dr. Death,' and the death-penalty machine that used him. FACT: over roughly three decades Grigson testified in 167 Texas capital trials, overwhelmingly for the prosecution, nearly all of which ended in death sentences; his value was predictability, telling juries plainly that the defendant was a sociopath certain to kill again. FACT: Texas capital sentencing requires a jury to predict 'future dangerousness' before imposing death, and Grigson supplied that prediction with a confidence no honest clinician could offer — often '100 percent' certainty, and frequently without examining the defendant at all, reasoning instead from a prosecutor's hypothetical; the field does not support such certainty, and in the 1983 Supreme Court case Barefoot v. Estelle the American Psychiatric Association told the Court that psychiatric predictions of long-term future violence are wrong more often than they are right. FACT: Grigson helped send Randall Dale Adams to death row for the 1976 murder of Dallas police officer Robert Wood — a crime Adams did not commit (the real shooter was teenager David Harris) — a wrongful conviction unraveled by Errol Morris's 1988 documentary The Thin Blue Line, which Morris had originally set out to make about Grigson himself; Adams was exonerated in 1989 after twelve years in prison. FACT: in 1995 the American Psychiatric Association and the Texas Society of Psychiatric Physicians expelled Grigson, on the grounds that he diagnosed people he had never examined and claimed he could predict future violence with 100 percent certainty; Grigson blamed 'anti-death-penalty doctors,' but the charge was ethical, not political. FACT: his expulsion did not undo the verdicts — Texas courts continued to uphold death sentences resting on his testimony even after his discrediting, because junk science, once admitted and acted on, is extraordinarily hard to pull back out of a verdict. THE HONEST FENCE: this is not a claim that everyone Grigson testified against was innocent — most were not, and guilt of the crime is a different question from a reliable prediction of future violence; it is not an attack on psychiatry, since his own profession's associations expelled him for violating its ethics; and it is not a claim that he single-handedly caused each sentence, since prosecutors chose him, judges admitted him, and juries believed him. The structural point is that a capital-sentencing system built on predicting the future created a market for a witness willing to fake the certainty, that the market was filled 167 times, and that at least once it helped condemn an innocent man. The hardest question: is the problem the man who faked the prophecy, or a death penalty that made a prophecy a prerequisite for killing?
Trial by Junk Science
The broader forensic-junk-science reckoning, generalizing the Dr. Grigson case, and graded entirely from the institutions' own findings. FACT: the National Academy of Sciences' 2009 report 'Strengthening Forensic Science in the United States: A Path Forward' concluded that, with the sole exception of nuclear DNA analysis, no forensic method had been rigorously shown to reliably connect a piece of evidence to a specific individual or source; the 2016 President's Council of Advisors on Science and Technology (PCAST) report confirmed that the leading feature-comparison methods — bite marks, firearms toolmarks, footwear, microscopic hair — lacked the 'foundational validity' needed for the identification claims examiners made in court. FACT: in 2015 the FBI and Justice Department admitted that, for decades before 2000, nearly every examiner in the FBI's microscopic hair comparison unit had given scientifically flawed testimony overstating hair 'matches' — a review of 268 trials found 26 of the 28 examiners made erroneous statements favoring the prosecution, errors in at least 90 percent of cases reviewed, including the cases of 32 people sentenced to death; the review was triggered by exonerations such as Santae Tribble, who served 28 years on a hair 'match' DNA later disproved. FACT: bite-mark comparison has no established scientific basis (per both NAS and PCAST) yet convicted people for decades — Keith Harward spent 33 years in prison on bite-mark testimony before DNA exonerated him in 2016, one of more than two dozen known bite-mark wrongful convictions or charges. FACT: the old arson 'indicators' were largely folklore rather than validated science, and Cameron Todd Willingham was executed by Texas in 2004 for a 1991 fire that fire scientist Gerald Hurst (before the execution) and Craig Beyler's report for the Texas Forensic Science Commission (after) found showed no scientific basis for a finding of arson — we grade FACT that the arson evidence was invalid, while noting his innocence is, by the nature of an execution, no longer provable. FACT: DNA analysis is the rigorously validated exception that exposed the rest, and false or misleading forensic evidence is a documented contributing factor in roughly a quarter of known exonerations. THE HONEST FENCE: this is not a claim that all forensic science is junk (DNA is real science and the great exonerator), nor that everyone convicted with these methods is innocent (the wrong is the overstated certainty and the lack of validation, not blanket innocence), nor that all examiners are frauds (many sincerely believed their fields were sound); and reform since 2009 is genuine but partial and resisted, with the Justice Department publicly disputing PCAST's conclusions. The pattern is the Grigson move generalized — false certainty delivered in a lab coat, before a courtroom that rewarded confidence over candor and rarely checked the science. The hardest question: if the FBI's own review found its hair examiners wrong in 90 percent of cases, including 32 death sentences, how many of those verdicts have actually been overturned — and what does the answer say about whether the system wants to know?
The Benefit of the Doubt
A graded account of the October 2026 killing of Koran Irvine on the George Washington Bridge, built as the two-tier-justice question it raises rather than an assertion of a fix. FACT: on the night of October 20, 2026, Koran Irvine — a 46-year-old Con Edison worker and married father of two from Valley Cottage, New York — was killed in what prosecutors describe as a road-rage shooting; according to the criminal complaint and surveillance video, Victor Romero, a 62-year-old who retired from the NYPD about 25 years ago, tailgated Irvine after a near-collision on a bridge entrance ramp, pulled alongside in moving traffic, and fired a single shot that struck Irvine in the right shoulder, after which Irvine's car crashed and he was found dead and Romero drove off toward home; surveillance placed Romero alone in his vehicle, and investigators recovered a registered revolver and two more handguns. FACT: the NYPD initially filed a murder charge, the district attorney's complaint instead charged first-degree manslaughter, Romero was held on $1 million bail, and he posted it and was released; Irvine's family, present at the arraignment, condemned both the downgrade and the release and said they would commission an independent autopsy. FACT (mechanics): the release follows structurally from the charge — under New York law second-degree murder is a qualifying offense for which a defendant can be held without bail while first-degree manslaughter is bail-eligible, so the decisive step is the charge reduction, where prosecutorial discretion lives, not the bail ruling itself. SOME SMOKE (posed, not asserted): the question of whether a former officer's law-enforcement identity earned him a benefit of the doubt a civilian would not get — the defense opened by calling Romero a 'hero cop' and a family man and telling the court 'this man is not going anywhere,' and there is a documented backdrop of 'professional courtesy' and systemic deference to law-enforcement defendants, but we fence it hard because the murder-to-manslaughter downgrade has a legitimate evidentiary basis unrelated to any badge (a single shot fired in a road-rage flashpoint makes the specific intent to kill that murder requires genuinely hard to prove, and manslaughter is the charge built for reckless or intent-to-injure conduct), and there is no evidence on the current record that the DA extended Romero a favor. THE HONEST FENCE: Romero is charged, not convicted, and everything here is allegation and the prosecution's account; the reduced charge is a routine and defensible reading of these facts; and this is not an on-duty police shooting, so the qualified-immunity and grand-jury dynamics that shield working officers do not apply to a man who left the force a quarter-century ago. What survives is a man dead over a moment of traffic, his alleged killer home within days, and a defense that led with the badge — a live public test of whether the outcome tracks the evidence or the résumé. The hardest question: if a civilian with no record were on video tailing a stranger across a bridge and firing into his car, would the murder charge have survived to arraignment — and if the honest answer is 'maybe not,' what is the word for a system where that's true?
A workplace injury: the FECA doctrine that classified an alleged rape as an on-the-job hazard
Elsie Dominguez, a civilian 1st Assistant Engineer aboard the USNS Carson City, sued the United States in November 2023 alleging the Navy's negligence enabled the ship's captain to drug and rape her in December 2021. In August 2026 the case was dismissed - not on the facts but on a doctrine. Because she lived aboard in temporary travel status, the Department of Labor determined her claim was a compensable workplace injury under the Federal Employees' Compensation Act, whose benefits are the exclusive remedy (Johansen v. United States, 343 U.S. 427 (1952)), so the court lacked jurisdiction to hear a civil suit. The legal machinery - assault classed as a workplace injury, comp as the only remedy - is graded FACT and is the story. The assault itself is graded as an allegation throughout: Captain Yamil Sanchez Padilla was never criminally convicted and the criminal case was withdrawn. The doctrine is live and reaches every federal worker lodged on the employer's premises in travel status.
Questions worth taking seriously
Aren't you just saying judges and prosecutors are corrupt?
No. The large majority do their jobs within the rules, and we say so. The claim is narrower and better documented: when misconduct does occur and is proven, the system publicly disciplines almost no one, resolves most of it in secret, and — for prosecutors — legally forbids the victim from suing. It’s a structural accountability failure, not a claim that everyone on the bench or at the prosecution table is dirty.
Doesn't absolute immunity exist for good reasons?
The Supreme Court’s stated reasons are real — preventing a flood of retaliatory suits, and not chilling tough charging decisions. We lay those out fairly. But they don’t explain why immunity should cover a prosecutor who knowingly fabricates evidence or uses perjury, and the Court’s fallback — that bars and elections will police misconduct — collides with the data: a 2013 study found fewer than 2% of prosecutors in thousands of misconduct cases faced any public discipline. That gap is the open question the immunity spoke presses.
If you are named on this page
If you are named on this page, or are a party materially affected by the claims made here, and you wish to respond, correct the record, or add context, use the Contact page. Responses are published verbatim alongside the original claim, with the sender identified and the date of receipt. The channel stays open for the life of the page.
This site aggregates and grades a record that other outlets and primary sources have already put on the record. Every FACT-graded claim above is sourced to court filings, government reports, sworn whistleblower disclosures, published investigative journalism, or named-source statements. The citations are the accountability mechanism; this section is how you get on the record too.