The Massachusetts Supreme Court earlier this year interjected itself again in the case of Victor Rosario, a man who spent 32 years in prison for a deadly fire. The justices in May refused to send Rosario back to prison and refused, too, to overturn a trial judge’s 2014 ruling that had set him free in the first place because of new doubts about the 1983 arson and murder conviction against him.
This sort of result is not uncommon in Massachusetts or around the country. New evidence is uncovered. Old witnesses recant. Prosecutors are found to have cheated. Another person confesses to the crime. Technology changes. A murder weapon is discovered with someone else’s prints on it. There are plenty of ways old convictions become “wrongful” over time. But the results of the Rosario case are notable for at least three reasons.
Rosario was convicted based in part on a confession, but the high court accepted a late-arriving medical diagnosis of “delirium tremens” (drug or alcohol withdrawal, the “DTs”) to undermine the credibility of that confession. He was convicted based in part on forensics, but the high court applied new fire science to undermine the reliability of the prosecution’s account. Finally, both the trial judge and the state supreme court applied a standard of review that if applied more often would make it much easier for defendants everywhere to get their convictions overturned.
In most cases where convicted defendants seek to overturn a guilty verdict, they must convince judges both that substantial new evidence was discovered after trial and that the introduction of that evidence would have made a difference in the outcome of the trial. Even if defendants can meet the first part of that test, they almost always fail the second part. If prosecutors withheld evidence, for example, the reviewing court will point to a confession and conclude the misconduct was immaterial. Judges do this, they say, in the interests of “finality” and “certainty” in criminal cases.
Not in the Rosario case. Not in Massachusetts. The justices tied all of the problems together and concluded that as a whole his conviction created “a substantial risk of a miscarriage of justice.” New doubts about the accuracy of the fire science, for example, would have raised doubts for jurors about the reliability of Rosario’s confession. And doubts about the reliability of the confession, in turn, prompted judges to conclude that trial jurors would have been far more skeptical of the fire science presented to them.
Prosecutors are expected to announce in a few weeks whether they intend to retry Rosario. If they do proceed it’s difficult to see how they’ll gain another conviction. The witnesses against him have not gotten more convincing over time. The forensic evidence that was so critical at his trial has been discredited. And we now know the extent to which Rosario’s brain was addled as he detoxed in an interrogation room with the police suggesting theories of the crime to him.
Then there is Rosario himself. After he went to prison, after he stopped drinking, he found God. Or maybe God found him. He became one of the first ministers trained and then ordained in a Massachusetts prison. Since his release in 2014 he’s preached at church services all across the Commonwealth. He’s voted, drives, runs marathons, and got married to a woman he met while he was behind bars. If prosecutors want to try him because they believe he’s a danger to society, or because they feel he’s beyond redemption, they’ll have a tough time justifying those conclusions to a jury.
It’s been 35 years since a fire fully engulfed a multi-unit, multi-story apartment building in Lowell, Massachusetts. The deputy chief of the city’s fire department later said he could see the glow of the fire as he left the firehouse early in the morning of March 5, 1982. By the time the flames were extinguished,eight people had perished inside the building. Men, women, and children, all dead from carbon monoxide poisoning and smoke inhalation.
The fire and its toll were big news all over Massachusetts. The fire’s rapid spread and burn marks immediately drew the attention of arson investigators. They became convinced that the fire was not accidental and that it could have been started by “Molotov cocktails” thrown through windows on the first floor of the building. One witness to the fire said he saw three men standing outside the building before the fire started, that he heard the sound of breaking glass and one man with his arm raised, as if he were throwing or had thrown something.
There was no doubt that Victor Rosario was at the scene of the fire; a Red Cross worker treated him there for a cut on his hand and sent him to the hospital. Another witness said Rosario had been at her apartment across the street doing drugs on the night of the fire and that she had later seen him breaking windows. But there were questions even then about the accuracy of those identifications. The witness who saw those three men standing outside the building twice failed to pick out Rosario from a photo lineup at the police station; only after Rosario’s photograph appeared in local newspapers did he incriminate Rosario.
The police arrested Rosario and questioned him on the night of March 6 into the following morning. Because English was not his first language Rosario was given a translator. Over the course of an eight-hour interrogation, during which Rosario showed signs of incoherence, he signed a series of three incriminating statements to the police, confessing to starting the fire along with two other men by throwing Molotov cocktails into the building. Rosario even helpfully offered a motive; one of other men wanted revenge for a drug deal gone bad. And then, as soon as he was done confessing, he claimed that he was “the son of God” and that he believed the back of his head had been cut off. His condition became so severe that morning that he was diagnosed as “psychotic” by doctors at the state mental hospital.
Rosario was charged with arson and murder. Prosecutors relied on forensic arson experts to educate jurors about the cause of the fire. They relied on the eyewitnesses who had linked Rosario to the crime scene. And then they used Rosario’s confession to wrap it all up. The defense, meanwhile, told jurors that Rosario had cut his hand that night trying to rescue children from the building and that Rosario, a heavy drinker and drug user, was psychotic at the time of his confession. Rosario testified in his own defense, telling jurors that he was near that building trying to save lives, not take them. Jurors quickly convicted him. He was given eight concurrent life sentences.
At Rosario’s trial fire investigators told jurors they were sure the fire had been deliberately set in at least two places with a “liquid accelerant.” They reached this conclusion in part because some areas of the building endured heavy burning and smoke while other areas “had no burning whatsoever.” They concluded that the “amount of fire coming from different isolated locations” also suggested that the fire was not accidental. The presentation at trial must have been quite impressive; it was decades before the fire science that was introduced in so many arson cases of that time were thoroughly discredited.
Rosario began challenging the expert testimony used against him shortly after he was convicted. In 1985 he argued that he should have been allowed to argue that there was a pattern of fires in buildings with common ownership to the one in the case. A judge quickly rejected that argument. In 1994 and 1995, Rosario came back to court again to challenge his conviction. This time he argued it was flawed because no flammable fluids were detected at the fire scene and because the trial judge had allowed his confession to be read to jurors. The same judge quickly rejected these arguments as well.
Nearly 20 years later, in 2012, Rosario was back again. This time he had science on his side. Two fire experts hired by a new attorney assigned to the case by the Massachusetts Committee for Public Counsel Services Innocence Program concluded that the fire that night in Lowell was, as the court later put it, “equally susceptible to an interpretation that the fire was accidental, involved no flammable liquids, and had a single point of origin.” The experts explained to the trial judge in a subsequent hearing that the phenomenon of “flashover” likely took place. It occurs when a fire leaps from burning fuel in a contained space (like a room) to the oxygen available in the whole area of the fire. “Flashover,” these experts told the judge, explained how the fire travelled that night.
Rosario’s team also attacked anew the reliability of the confessions he had signed in 1982. At trial his lawyers told jurors that their client was a drunk, with a history of drug and alcohol abuse and a prior head injury that made him even more susceptible to the ravages of his lifestyle. Defense witnesses also told jurors that Rosario had been drinking heavily in the days preceding the fire but that he had “dramatically” stopped drinking shortly after the fire. There was also the testimony that he had been psychotic during the interrogation. None of this moved the jury, of course, but it did provide subsequent experts with a timeline that ultimately helped Rosario.
He had not been psychotic while the police were interrogating him, defense psychiatrists told the trial judge during the 2014 hearing in the case. He was suffering instead from an acute case of drug and alcohol withdrawal exacerbated by his prior head injury. When he arrived at the police station for the interrogation that would lead to his confession, he was finishing his second day of withdrawal, at a time when his withdrawal symptoms would be fully ripe. Moreover, the translator that night, whose credibility is not in question, later testified that Rosario did not actually volunteer the incriminating information. The officers interrogating him did. To the state supreme court, and to the trial judge before it, the new diagnosis combined with the new context about the interrogation tactics the police employed justified overturning Rosario’s conviction.
“There are substantial differences between psychosis and DTs that may have made a real difference in the jury’s verdict,” the justices wrote in their May ruling. A person in the throes of withdrawal, the justices said, could not have the state of mind necessary to make a confession deemed “voluntary” under the law. It’s not just a ruling that helps Rosario stay free pending the decision to re-try him. It’s also a ruling that will make it harder for prosecutors to use his “confession” again to convict him if he is retried.
There have been countless cases all over the country where defendants have introduced compelling new evidence only to be denied a new trial or a new sentence because reviewing judges have concluded that the new information would not have convinced jurors to vote for acquittal. In these instances, judges typically look at each new issue raised and evaluate whether it, alone, would have altered the jury’s verdict. This standard makes it more likely judges will reject the appeal. What the courts in Massachusetts did in this case, however, was to “look beyond the specific, individual reasons for granting a new trial to consider how a number of factors act in concert to cause a substantial risk of a miscarriage of justice,” as the justices put it.
A chain is only as strong as its weakest link, the judges seemed to say. Doubts about the fire science would have undermined the weight jurors gave to the reliability of his confession. And doubts about his confession would have undermined the weight jurors gave to the accuracy of the forensic science employed against him at trial. Both of those factors would have cast doubt on the accuracy of the eyewitness testimony.
It is impossible to say how many more convictions would be overturned on appeal if more courts around the country employed this standard, which focuses on the totality of the evidence, rather than the more restrictive standard, which emphasizes the need for cases to have a definitive end even if that end undermines confidence in the reliability of a conviction. Illinois, Kentucky, and Connecticut also employ a form of what I’ll call the “Weakest Link Standard.” But in many jurisdictions, Rosario’s appeals wouldn’t have gotten him a hearing, much less a chance to walk free.