The state's case against Shirley Ree Smith wasn't the greatest. Etzel Glass was only seven weeks old when he died, but arguing that Smith had shaken him to death was a stretch. She was helping her daughter raise Etzel and two other children, and there was nothing to indicate she'd been anything other than loving toward them, and none of the nonmedical evidence that normally arises in shaken baby cases: a past pattern of abuse, an unwanted child, a "trigger" such as inability to stop the baby from crying. The medical evidence wasn't any great shakes, either: no evidence of physical injury, no retinal bleeding or brain swelling, and only minimal subdural hemorrhaging. In fact, the state's argument was basically that the child had died from shaking which had sheared his brain stem in a manner which couldn't be detected.
But this was in 1997, the heyday of shaken baby syndrome. It wasn't unusual to drive by a billboard admonishing young parents not to shake their babies, and it was the same year that Louise Woodward, the English nanny, was convicted in a notorious case of shaking her 8-month-old charge to death. The jury found the evidence against Smith sufficient for conviction, so off she went to do a 15-to-life stretch.
Nine years later, the 9th Circuit reversed, finding that the evidence was clearly insufficient to support the conviction. In the first decision of this term, the Supreme Court, in a per curiam opinion, reversed that, and sent Smith back to prison. But, as Justice Ginsburg's opinion for the three dissenters explains, there's more to the story than that.
First, there's the subtext of why the Court took the case in the first place. Ginsburg notes that "error correction" has long been viewed to be "outside the mainstream of this Court's functions"; the Court's own rules provide that "a petition for certiorari is rarely granted when the asserted error is the misapplication of a properly stated rule of law." The 9th Circuit based its decision on Jackson v. Virginia, the Supreme Court case holding that a conviction obtained on insufficient evidence violates a defendant's due process rights. Maybe the 9th got it right, maybe it got it wrong, but what's so extraordinary about the case that it warrants being one of the 80 cases the Court will be asked to hear, out of thousands and thousands of petitions, let alone be resolved in such summary fashion, without full briefing or oral argument?
If you follow the Supreme Court at all, you know the two most important words in the preceding paragraph are "9th Circuit." When you have the most liberal court in the country butting heads with the most conservative Supreme Court in modern history, it's sort of like the old Godzilla/Megatron clashes you'd see on late-night TV, except there's no doubt as to the outcome. Cases from the 9th Circuit make up a staggering percentage of the Court's docket -- a third of it last term, and nearly 40% this term -- and the results aren't pretty; when the dust settled last year, for example, the Court had reversed the 9th in 19 out of 26 cases.
And at least at cursory glance, it's not hard to fault the Court in this instance. After all, this was a habeas case, with an exceedingly deferential standard of review toward the state courts' findings, all of which had rejected the insufficiency claims (as had the magistrate and the judge in the Federal district court.) What's more, the Court had twice earlier vacated the 9th's decision and remanded the case for reconsideration in light of two other Supreme Court decisions, and the 9th had failed to take the hint either time.
But Ginsburg's opinion points to another issue: what had once been an undisputed diagnosis is now a subject of controversy. For example, in 2003 the American Journal of Forensic Medicine and Pathol0gy did a review of the 30-year literature on the syndrome, and determined that "there was inadequate scientific evidence to come to a firm conclusion on most aspects of causation, diagnosis, treatment, or any other matterspertaining to SBS." There are now questions as to whether the forces necessary to cause damage to an infant's brain can be done without causing injuries to the cervical spine or brain stem; no such injuries were found in Etzel. Indeed, Ginsburg, quoting from a later article in the same journal, notes that some experts now argued that the entire theory of how shaken baby injuries occur is "contrary to the laws of injury biomechanics as they apply specifically to the infant anatomy."
That may well be overstating the situation; many doctors hotly contest that claim. But the point is that there is tremendous controversy over this where there was none before. Even the pediatric neurosurgeon who came up with the diagnosis in 1971 is having some second thoughts about its application, prompted by this New York Times article highlighting the growing dispute over the validity of the diagnosis in numerous cases.
Does that mean that Shirley Ree Smith was twice the victim -- the injustice of her original conviction, compounded by being caught in the crossfire between the 9th Circuit and the Supreme Court? I don't know, but California Governor Brown has his thoughts on the issue: two weeks ago, on Good Friday, he commuted Smith's sentence to time served.