One for two
If you happen to be suffering from too much self-esteem, here's the remedy: start handling Federal habeas corpus cases. There were two decisions out of the 6th Circuit within the past couple of weeks. McCarley v. Kelly presents a nice Crawford issue, and shows that you can win these cases. Williams v. Bauman shows just how hard it is to do that.
We'll forget about the procedural hurdles for habeas relief, most of them created by Congress' passage of the Anti-terrorism and Effective Death Penalty Act in 1996. One of the major revisions was the requirement that a defendant have exhausted his state remedies. Since only Federal constitutional issues can be presented in habeas, that means the defendant has to give the state courts a shot at deciding those issues, to "fairly present" those issues. If you're handling appeals, and you're not framing your assignments of error wherever possible with some Federal constitutional violation, start.
But that just gets you over the first hump. Let's take a look at Williams for the problems in store after that. Williams and his co-defendant Coleman were charged with the robbery and murder of a shop-owner, and the key eyewitness, Banks, died before their joint trial. Without objection, his testimony in the preliminary hearings for both defendants was read into evidence.
Williams didn't have an opportunity to cross-examine Banks at Coleman's preliminary hearing, so he claims that his right to confrontation, and that his lawyer was ineffective for not objecting to it. The state court held this wasn't error, and decided that even if it was, it was harmless. The district court did find a violation, but agreed on the harmlessness issue.
That's where the fun starts. The AEDPA allows a writ to be granted only where the state court's decision on the merits was "contrary to, or involved an application of, clearly established Federal law," as determined by the US Supreme Court, or that the decision was "an unreasonable determination of the facts." The Supreme Court says this doesn't include error correction; habeas relief should be granted only where there is "an extreme malfunction" in the state's criminal justice process. It's a pretty high bar: "an unreasonable application of the Supreme Court's holdings must be objectively unreasonable, not merely wrong; even clear error will not suffice."
That's abuse of discretion on steroids.
Williams first argues that Banks' testimony in Williams' own preliminary hearing should have been barred, on the grounds that he didn't have an adequate opportunity to cross-examine. That's probably true. Back when they were doing preliminary hearings in this county, if you spent more than five minutes cross-examining someone, the judge would shut you down. There's actually some case law from the 6th Circuit itself which suggests that it might not be sufficient opportunity, but the Supreme Court hasn't ruled on the issue, so dooms Coleman's first contention.
Williams' second argument, that the judge erred in allowing Banks' testimony at Coleman's preliminary hearing to come in, fares better, at least for a moment or two; the court agrees that it's "difficult to square" the state court's allowance of this with Crawford and other Supreme Court cases.
But only for a moment or two, because the court moves to the issue of whether the error is harmless.
Well, that's got to be good news for Williams. After all, even a novice criminal lawyer knows that the standard of review for constitutional error is whether it's harmless beyond a reasonable doubt.
But not in habeas: the test is whether the state court's determination that the constitutional error was harmless is unreasonable; "an error requires reversal only if it had substantial and injurious effect or influence in determining the jury's verdict."
I'm not even sure what that means, but it's not good. The court decides it wasn't; there wasn't any real difference in Banks' testimony at the two preliminary hearings.
But Willard McCarley does hit the jackpot, in another Crawford case. The short version is that a 3-year-old boy witnessed his mother's murder. The police had him see a psychologist to try to get additional information about what he saw. Some ten years later, McCarley was charged with the crime, and the court allowed the psychologist to read the statements the boy made to her; the boy, now fourteen, didn't remember making the statements.
And here we learn a little trick. Federal courts give great deference to the state court's decision, but that applies only to the last "reasoned opinion" of the state court addressing the issue. That means the court of appeals' opinion, because the Ohio Supreme Court declined review. But the state appellate court's decision didn't resolve the issue of whether there had been a Confrontation Clause violation; it held that even assuming the admission of the psychologist's testimony was improper, the error was harmless.
So here's the trick: the 6th Circuit holds that the appeals court "did not issue a decision as to whether the State violated McCarley's Sixth Amendment rights; rather, it made a point of not deciding the issue."
What's that mean? Since there's no state court decision on the merits, there's no deference, and the Federal court reviews the issue de novo. I'd much rather have three Federal judges deciding whether the admission of the evidence was wrong than deciding whether the state court judges had gone so far off the reservation that they jumped the AEDPA hurdle.
The 6th Circuit decides that it was error, and then works through the harmlessness issue, coming out in a different place from Williams, and understandably so: The psychologists' testimony, on the other hand, was central to the prosecution's case.
One other trick here: if you've got a harmless error issue, look at the prosecutor's closing argument to see whether they highlight evidence that shouldn't have been admitted. If they're telling the jury that the evidence is important, it's going to be a lot more difficult for them to tell three appellate judges that it wasn't.
I remember running across a study of 1700 or so habeas cases, which found that 7 of the petitions had been granted. That's close to everything else I've heard, at least for non-capital cases. Going one for two in two weeks is unusual.