Why wrongful convictions are so hard to undo once prosecutors commit
Josh Dubin's account of the Ohio 4, a former prosecutor's explanation of why his colleagues cannot admit mistakes, and Harvey Silverglate's criticism of the FBI and political prosecutions all describe a system with few incentives to correct itself. The pattern matters because a person's freedom can depend less on the evidence than on whether an official is willing to reverse course.
A wrongful conviction is hard to undo less because the evidence still supports it than because nobody with the power to undo it gains anything by doing so. Three conversations describe the same pull toward charging and holding, and away from letting go. This is mostly one county's story, told by the defense lawyer and a prosecutor who came to agree with him, and the convictions in question still stand.
Josh Dubin, a civil rights attorney, describes four men convicted of murder in Ohio in the early 1990s. A reward was offered in a cold case, and the conviction came to rest on one witness who wanted money. There was no forensic evidence and no other eyewitness, and a federal agency had stopped using the witness for taking reward money for false information. He later recanted more than once. A federal appeals court found a credible claim of innocence for one of the men. But when the witness was called, he was warned about perjury, refused immunity, and declined to testify. Dubin says courts view recantations with deep suspicion. His summary is "it's really easy to put someone in jail, real easy, real difficult to get them out".
J.D. Tomlinson, the former prosecutor of that county, reviewed the record looking for any evidence of guilt and says he never found it. Before leaving office he joined the defense in asking for new trials. The judge did not hold the usual hearing, and his successor withdrew the state's support on his second day. Tomlinson had earlier exonerated two other people and apologized to them in court, after which police and former prosecutors turned on him. He says prosecutors want to win. He does not claim most convictions are wrong, and says most of his office's cases are probably good.
Harvey Silverglate, a civil liberties lawyer, describes the federal version. Agents decide who is guilty and then build the case, he says, and their interview reports record what they wish you had said. His material is about investigations, so it supports the pattern only by analogy.
They agree that the commitment comes before the evidence and that publicity is one of the few levers. The mechanisms they name are specific. Winning is the job and careers depend on it. Dubin says "prosecutors have immunity There are no consequences" even for withholding evidence. A recanting witness is trapped by the threat of perjury. Ego and loyalty to a former office do the rest. They differ on whom to trust and on the fix: an outside review panel, accountability, defense experience for prosecutors, or replacing an agency.
What to do with it: vote for county prosecutor, an office Tomlinson says has more power over your life than the president. Read the record before taking a side. Use lawful public pressure. Ask what a conviction rests on, since a single paid witness with no forensic evidence is the warning profile. Back accountability for withheld evidence. Practice saying you were wrong. Anyone facing charges or seeking review needs a lawyer, and nothing here is legal advice.