In 1996, when he was seven weeks old, Etzel Glass died during the night. His mother, Tomeka, had put him to sleep on a sofa in a room with Shirley Ree Smith, Etzel’s grandmother. Smith was helping Tomeka raise Etzel and two other children, who were also sleeping in the room with her. There was no indication that she’d been anything but loving toward the kids at any time. When Smith woke up and found Etzel limp, she ran with him to his mother’s room next door, saying she thought he’d fallen off the sofa. At first, the doctors who examined Etzel said he’d died of sudden infant death syndrome—no one’s fault. But the coroner found the cause of death to be shaken-baby syndrome, and prosecutors decided that Shirley Ree Smith had done the shaking.
This was not a typical shaken-baby case. The standard diagnosis for shaken-baby syndrome includes subdural bleeding, retinal bleeding, and brain swelling. In the cases that are easier to classify, there is also injury to the neck from shaking, or there are fractures, bruises, or cuts. In the harder cases, there are no external signs of injury. Etzel’s case involved only “minimal” subdural hemorrhaging. There was no retinal bleeding and no brain swelling, and no fractures or abrasions.
But there was one more supposed piece of evidence against Shirley Ree Smith. According to the social worker who interviewed her, Smith said that when Etzel didn’t respond to her touch when she thought he was sleeping, she gave him “a little shake, a little jostle.” Smith then said something like “Oh my God. Did I do it? Did I do it? Oh my God.” This was enough for the prosecutors and the jury. She was convicted and sentenced to 15 years to life.
I’d like to think that Shirley Ree Smith would not be prosecuted today, though I’m not certain of it. The science underlying shaken-baby prosecutions is shifting, with critics questioning whether alternate explanations for a baby’s death are always adequately explored. But a new consensus—legal or scientific—hasn’t yet emerged yet from the bitter fight, in some cases, over the diagnosis.
Smith’s case is truly an outlier. Since Etzel didn’t have the typical symptoms, the prosecution’s medical witnesses went out on a limb, testifying that the death was caused by a shearing or tearing of the brain stem even though no doctor located any such shearing or tear. On appeal, Smith argued that there wasn’t enough evidence for a jury to find her guilty. But shaken-baby convictions aren’t easy to undo. For the California appeals court, it was enough that the prosecutors had medical experts on their side. The California Supreme Court declined to review the case. So Smith turned to the federal courts. At this point, the Anti-terrorism and Effective Death Penalty Act kicks in. This is perhaps my least-favorite law. AEDPA tells federal courts that they can’t overturn state courts except in a narrow set of circumstances: If a conviction is contrary to or unreasonably applies clearly established federal law, or if it’s based on an unreasonable determination of the facts. AEDPA is a straitjacket. The federal courts are supposed to put it on and quit worrying about whether innocent people have been put in prison.
A panel of judges for the U.S. Court of Appeals for the 9th Circuit refused to do that. They looked at the medical testimony against Shirley Ree Smith and how badly it holds up to the light of current knowledge, and they said that Smith had spent enough years in prison.
Normally, that would be that. AEDPA or no AEDPA, the Supreme Court doesn’t usually reach out and grab sad, fact-bound cases like this one. But liberal 9th Circuit judges are a thorn in the side of the Supreme Court’s conservative justices. Last month, after twice sending the 9th Circuit pointed warnings about this case, the Supreme Court reversed the circuit court’s decision. The majority’s brief and unsigned opinion concedes that “doubts about whether Smith is in fact guilty are understandable.” But according to six justices, it’s not the 9th Circuit’s job to do anything about that.
Justice Ruth Bader Ginsburg dissented, with Stephen Breyer and Sonia Sotomayor. Ginsburg gives all the reasons to doubt the medical testimony against Smith. She does a great service by laying out the growing skepticism among a minority of doctors about the validity of diagnosing shaken-baby syndrome without any evidence at all of external injury. “What is now known about the SBS hypotheses seems to me worthy of considerable weight in the discretionary decision whether to take up this tragic case,” Ginsburg writes.
As for the weight to accord Smith’s supposed confession, I defer to an amazing opinionpublished last week by Judge Richard Posner. In this case, a daycare provider named Rick Aleman was charged for shaking an 11-month-old baby who collapsed in his care. When the police told him that three doctors said that Aleman’s shaking must have caused the injury, he said “I know in my heart that if the only way to cause [the injuries] is to shake that baby, then, when I shook that baby, I hurt that baby.” It turned out that the police were lying. And that the baby’s mother had a record of violence and crime. And that the cop who lied apparently wanted to date her. And that the baby had been sluggish and lethargic for days before he came to Aleman’s house.
The case against Aleman unraveled before trial. It came to Posner and his fellow judges on the 7th Circuit as a suit, brought by Aleman, for false arrest and malicious prosecution. In sustaining those claims, Posner explains how easily an innocent person can implicate himself. The police told Aleman that the only possible explanation for Joshua’s injuries was that the baby had been shaken right before he collapsed. Aleman was the only person to have shaken Joshua just before he stopped breathing. “And so it was a logical necessity that he had been responsible for the child’s death. Q.E.D.,” Posner says. “A confession so induced is worthless as evidence, and as a premise for an arrest.”
That is exactly right. It also perfectly illustrates a serious problem in shaken-baby cases, where the accused often held or handled a baby before the awful moment of collapse, and may be easily led to blame himself.
In Shirley Ree Smith’s case, Justice Ginsburg concludes, “I would not ignore Smith’s plight and choose her case as a fit opportunity to teach the 9th Circuit a lesson.” That’s exactly right, too. There is only one lesson worth learning from this case, and it is about the power of mercy. California Gov. Jerry Brown has the authority to pardon Shirley Ree Smith. She has suffered more than enough for the death of her grandson. Brown should do for her what no judge now can.
x x x."
