Kind of a quirky published opinion yesterday from the Second Circuit.
But what's a little weirder about the opinion is that the court re-opens the question of what harmless error standard applies when the state court has engaged in its own harmless error analysis. In my mind, that question was settled by the Supreme Court a couple of years ago.
And then finally the court engages in an outcome-determinative harmless error analysis -- downgrading every type of harmless error analysis that looks at the impact of the evidence on the case, and basically equating harmless error analysis with solely looking at the strength of the prosecution's case. It's a little dispiriting. Try to explain to a layperson that a conviction can be upheld where there were two significant constitutional violations and the remaining evidence against the defendant consisted of circumstantial evidence and a shaky oral confession.
The case is Perkins v. Herbert, opinion here, which was written by DJ Trager. The court reversed the district court's grant of habeas.
Below the fold I'll go into a little more analysis . . .
Once again, the central question in the case is whether these two significant constitutional violations were harmless.
It's important to note right up front that both the state court and the district court concluded that there had been two significant constitutional violations. In the Second Circuit, the State asked the court to conclude that these other courts got it wrong and that there were no constitutional violations. The court entertains the argument, but rejects it.
I find it a little troubling that the court would entertain this argument. In the abstract, the Second Circuit should review that part of the case. At its core, habeas relief should only be granted if there actually were constitutional violations.
But thanks to the AEDPA, the logic is all messed up here. The standard of review says that habeas relief can only be granted if the state court's decision was either contrary to, or an unreasonable application of, clearly established federal law.
Here, the State was basically asking the Second Circuit to conclude that the state court's decision in petitioner's favor was an unreasonable application of federal law (or at least an incorrect one), and because of that, to deny relief.
But that doesn't fit into the standard, does it? According to the standard of review, when there has been an unreasoanble application of federal law, that's when habeas relief is granted. I am sorry, but, if Congress wants to restrict the federal court's power as it did with the AEDPA, then everyone needs to live with the consequences. And one of those consequences is that, when the state court issues a ruling in petitioner's favor, it's beyond the federal court's powers to ask if it's correct. Is it constitutional for Congress to do that? I don't know. But that's the consequence of the way that they screwed around with habeas.
Back to the case, after rejecting the State's argument, the court turns to harmless error. It states that the question of what harmless error analysis to apply here is an open question. Essentially, there are two different standards: (1) filtering the state's harmless error analysis through the standard of review and asking whether the state court's conclusion was unreasonable (what the court calls the AEDPA/Chapman test); (2) or the arguably more strict "actual prejudice" approach set forth in Brecht.
While I welcome the re-opening of this question as it could be favorable to petitioners, I don't think that it's an open question. I thought I had previously discussed all of this, but I can't seem to find it on the blog. I taught a CLE on it, though. Maybe I never posted the stuff up here. So if you have time, you can read through the following indented stuff that I wrote for the CLE. Otherwise, skip past it:
Prior to the AEDPA, the Supreme Court concluded in Brecht v. Abramson that the Kotteakos “substantial and injurious effect” standard applies on habeas. As we all know, this is a much tougher standard that the Chapman standard for constitutional violations.
After the AEDPA, the question was raised as to whether the standard of review alters the analysis? in other words, does a state’s harmless error analysis get filtered through the standard of review? Does Brecht apply no matter what? has the AEDPA overruled Brecht entirely and harmless error review reverts back to Chapman in certain situations? such as, what happens when the state court fails to engage in a harmless error analysis?
This ended up being a very tricky area. It was so tricky that, within the span of 4 years, the Supreme Court changed the answer to these questions (without even acknowledging it) and finally concluded that harmless error lies outside of the standard of review. The Brecht standard applies no matter what the state court did.
The first time the court faced the question was in 2003 in Mitchell v. Esparza.
In that case, the Sixth Circuit concluded that the state court had applied harmless error analysis in a situation where harmless error should not have applied.
The Supreme Court stated that the Sixth Circuit was wrong and harmless error was appropriate. And here’s the important part of the decision for this discussion. The Court then stated that what the Sixth Court should have done is apply the AEDPA. What the circuit court should have asked was whether the state court’s Chapman harmless error analysis was “unreasonable” under the 2254(d).
As can be seen, this was a reactive decision without much thought going into how the harmless error standard should apply in a habeas case. It’s more of a smackdown against the Sixth Circuit, the Court saying – You dummies, you must apply the AEDPA. See look how its done.
More important, the standard used here is a change from pre-AEDPA law. Rather than applying the Brecht standard, the Court applied, what it later described, as the more liberal Chapman/AEDPA analysis. That would appear to be favorable to habeas petitioners.
But the good times did not last long. Four years later, the Supreme Court addressed the question of what harmless error standard to use when the state court does not engage in a harmless error analysis.
Taking the opposite approach than the de novo vs. deference argument that a petitioner would usually use and relying upon Esparza, the petitioner argued that the AEDPA had overruled Brecht and all harmless error claims should fall under the Chapman/AEDPA standard.
Justice Scalia rejected this argument. He basically said that it is illogical to think that, in passing the AEDPA, Congress wanted to do anything to make it easier for a petitioner to obtain habeas relief. It’s an oppressive statute, not a liberating one. Those are my words. So in no way can the Supreme Court do anything to make habeas petitioner’s lives better. It is this reasoning why I think there was a concern that the de novo vs. deference question was going to come down the wrong way. The AEDPA was intended to restrict habeas relief, so how can courts fall back on pre-AEDPA standards to grant relief. But now they can.
Anyways, getting back to Fry, Scalia concluded the opinion by stating that, in a 2254 proceeding, the federal courts must use the Brecht standard whether or not the state appellate court recognized the error and reviewed it for harmlessness under the Chapman standard. Although not saying it, the Court just overruled Esparza and actually created an exception to the AEDPA standard of review. Talk about legislating from the bench.
The upshot is that in Fry, the Supreme Court stated, "We hold that in 2254 proceeedings a court must assess the prejudicial impact of constitutional error in a state-court criminal trial under the . . . standard set forth in Brecht[,] whether or not the state appellate court recognized the errer and reviewed it for harmlessness under the . . . standard set forth in Chapman . . . ." That is a holding of the Supreme Court that answers the question.
But as discussed in the indented material [I know, I suggested that you didn't have to read it; it's okay if you didn't], the Supreme Court had relied on the AEDPA/Chapman analysis in a prior case, Esparza, and did not specifically overrule that case in Fry. Based on those two things, the Second Circuit considers the question open. And the Second Circuit is not the only one. The Seventh Circuit also thinks that this is an open question.
I don't agree. While the Supreme Court should have specifically addressed the conflict between Esparza and Fry, this does not mean that Esparza is still good law. It isn't. The holding in Fry is pretty direct: it's the Brecht standard no matter what. It silently overruled Esparza. There is no unresolved question.
In the end, the Second Circuit here decides not to answer this "open question" and concludes that it was harmless under both tests.
And the court's harmless error analysis saddens me. Clearly, the court believes that this guy is guilty and does not want to give him another trial. As a result, the whole analysis feels completely outcome-determinative. I mean, these two erroneously admitted pieces of evidence represented the most important evidence in the case. They were precisely what the prosecution relied upon in summation to convict him. Certainly, petitioner was caught running from the scene and in possession of the victim's jewelry. That is strong circumstantial evidence. But circumstantial nonetheless. The court also places a lot of emphasis on petitioner's "oral confession." But this oral confession was the police telling petitioner what they believed happened and petitioner saying, "Yes, that's the way it happened." Not the most compelling confession.
And the worst part about the decision for habeas petitioners is how hard the court works to push aside every harmless error approach that looks at the impact of the evidence on the case. Those clearly were in petitioner's favor here. Rather, the court repeatedly emphasizes that, over the years, the single most important factor has become the strength of the remaining evidence. Okay, that's great. But do these other factors not matter at all?
I guess it really is a tug of war between a sense of fairness vs. a sense of practicality. It simply cannot be said that the trial was fair as a result of these two significant constitutional violations. But knowing everything that is known, there really wasn't too much of a question of guilt. The court essentially stated that, practically speaking, the guy's guilty, so we aren't going to force the State to go through the expense of a new trial. In doing so, significant questions about the fairness of the process got swept aside. Makes me sad.
Comments