Judge Merritt's Puzzling Dissent in Getsy v. Mitchell:
I agree with my co-blogger Jonathan that the en banc Sixth Circuit was correct in reversing the panel decision in Getsy v. Mitchell. I realize that different people will disagree about the death penalty. But as a matter of binding law, I don't see this case as even remotely close.

  In my view, the key to the case is that different trials with different lawyers and different juries will often reach different results. The Constitution clearly allows a defendant to be retried and convicted after a first jury hangs; there is no rule that the result on the second trial has to be the same as the result in the first one. Similarly, the Constitution allows one defendant to a conspiracy to receive a death sentence in one trial while another defendant receives a life sentence in another trial. I was quite puzzled by the contrary arguments put forward in Judge Merritt's dissent, and I thought I might blog a bit about them.

  Judge Merritt's main argument is that there is a common law English rule, adopted by the U.S. Supreme Court as a matter of Due Process in Morrison v. California, 291 U.S. 82 (1934), requiring consistent verdicts in conspiracy cases. If two individuals are tried together for conspiracy and the jury convicts one and acquits the other, the convicted defendant must be set free. Judge Merritt reasons that this same principle applies in cases with co-defendants tried separately, at least in death penalty cases, such that that Gentry can't get the death penalty

  I think there are a bunch of pretty serious problems with this argument. First, although this was the English common law rule, it turns out to be the minority rule in U.S jurisdictions today. As Eric Muller has explained, this rule is "now followed in only a few jurisdictions," and "[t]o date, almost every federal court of appeals has either abandoned or severely criticized the rule." The Hobgoblin of Little Minds? Our Foolish Law of Inconsistent Verdicts, 111 Harv. L. Rev. 771 (1998).

  As Eric notes in his article, the one circuit that seems to still follow the rule happens to be the Sixth Circuit. But that shouldn't matter given the standard of review in this case: Under AEDPA, the issue is whether the state court unreasonably applied clearly established Supreme Court precedent, not Sixth Circuit law or law that the Supreme Court might establish some day. Given that, I find it hard to see the relevance of the English common law rule.

  Further, even the Sixth Circuit has recognized that this so-called "rule of consistency" does not apply to separate trials. Here's what the Sixth Circuit said in United States v. Sachs, 801 F.2d 839, 845 (6th Cir. 1986), an opinion joined by Judge Merritt himself:
[I]f coconspirators are tried separately, the acquittal of all other coconspirators does not mandate acquittal as to the remaining conspirator. United States v. Roark, 753 F.2d 991 (11th Cir. 1985). This result is necessary because different juries may hear different evidence; "[t]hat the evidence was insufficient to support a guilty verdict in the one case does not mean that conviction on different evidence in another case was improper." Id. at 996. In other words, it is not necessarily inconsistent for two juries to reach differing results.
  As best I can tell, Judge Merritt has two responses. First, he says that it would be really unfair to not apply the rule in capital cases with separate trials because it would take away a legal defense from capital defendants. Capital defendants are given the extra process of having separate sentencing trials; under the Sachs rule, they will never be able to assert the defense of inconsistent verdicts. According to Judge Merritt, placing defendants in a setting where they cannot assert the defense of inconsistent verdicts "guts" the protections of the rule. If I understand the dissent correctly, Judge Merritt's answer is to expand the rule to make sure defendants can still make the claim even when they are tried in different trials.

  Maybe I'm just missing something, but I find this rather strange. The Constitution gives defendants the right to be free from inconsistent verdicts. It doesn't given them a right to assert a valid claim regardless of whether the facts support it.

  Finally, Judge Merritt places tremendous emphasis on Morrison v. California, 291 U.S. 82 (1934). Judge Merritt asserts that Morrison clearly and unambiguously constitutionally mandated the English common law rule that verdicts in conspiracy cases must be consistent — whether the case involves multiple defendants or not. The constitutionalization of the English rule is so clear, Judge Merritt claims, that it amounts to clearly established U.S. Supreme Court law. But an review of Morrison shows that Morrison does not clearly establish such a rule (as you might guess from the fact that most jurisdictions have rejected it without imagining any constitutional problems). As best I can tell, the relevant passage of Morrison seems to be about the common law requirements of conspiracy rather than the Due Process clause.

John C (mail):
Maybe a bit off topic, but . . .

Prof. Kerr seems right - it looks like Judge Merritt bungled this one. But doesn't this case point out the real problem with AEDPA's "whether the state court unreasonably applied clearly established Supreme Court precedent," requirement? I've always thought of that as a ridiculous command, putting courts in the position of trying to figure out what is "clearly established" and what is an "unreasonable application" of that precedent, both of which seem ripe for instrumentalism given the lack of any coherent method for figuring it out what counts and what doesn't.
7.26.2007 10:35am
John C,

There are pros and cons to AEDPA, but I personally don't find the idea of deferential review like that to be "ridiculous." Appellate courts are used to deferential standards of review, whether of facts ("substantial evidence" supporting a verdict) or of law (whether an agency's interpretation of an ambiguous statute is "reasonable" under Chevron). Such judgment calls can be hard, but that's why we call them judges.
7.26.2007 11:02am
Justin (mail):
Chevron is completely different. It involves deference - giving the agency lattitude to interpret the law, reviewed only under a particular set of standards. This is different - there is no allegation that the state law's action under AEDPA is controlling or even correct. The better anlogy appears to be Saucier v. Katz.
7.26.2007 11:28am

QI is also a good example, although I don't understand why Chevron is "completely different." As I see it, the issue in both Chevron and AEDPA is whether another decisionmaker's assessment of the law is either clearly wrong or, alternatively, at least possible.
7.26.2007 11:41am
Justin (mail):
OrinKerr, the reason for reviewing in Chevron is different - it's no accident that in Chevron, the deferential review only goes to the specialty of the agency, not to general principles of law - and thus the review is limited and bifuricated. In Chevron "step one," moreover, there is no deference at all. And the results are different - when the DC Circuit upholds an agency decision, it is announcing, in effect, that this is "the law," whereas in QI and AEDPA, they're just announcing that the actor in question didn't do anything too outrageous.

These differences also (should) effect the way a judge approaches a particular subject - and certainly impacts whether and how these approaches are constitutional/wise or not (although the arguments that any of it is unconstitutional is limited).
7.26.2007 11:47am
Justin (mail):
Or imagine a Kozinski of the left, writing one of his conversational dissent, using Saucier v. Katz analogy?

Lawyer: I have good news and bad news.

Defendant: What's the bad news?

Lawyer: You were unconstitutionally deprived of your rights.

Defendant: What's the good news.

Lawyer: The people who unconstitutionally deprived you of your rights didn't make a horrible mistake, and weren't malicious about it.

Defendant: Oh, good. So that means they'll correct the error? That means I get fair trial?

Lawyer: Well........
7.26.2007 11:51am

Obviously we're pretty far afield from the thread, but FWIW I tend to disagree.

First, Chevron Step One and Step Two collapse into each other in practice. Necause a slightly wrong interpretation of a very clear rule is unreasonable, the two step doctrine is indistinguishable in result from a general reasonableness rule. In fact, an empirical study found that in published decisions of the federal courts of appeals, courts reduced Chevron to a general requirement of reasonableness in 28% of the cases. See Orin S. Kerr, SHEDDING LIGHT ON CHEVRON: AN EMPIRICAL STUDY OF THE CHEVRON DOCTRINE IN THE U.S. COURTS OF APPEALS, 15 Yale J. on Reg. 1, 15 (1998). incidentally, the study found that the rates at which courts upheld agency interpretations when the test was phrased as simple reasonableness was the same as the rate when the inquiry was directly broken down into two steps. Id.

Second, if a court upholds an agency's view under Chevron, it doesn't mean it's "the law." It means it is the agency's interpretation until the decides to change it.
7.26.2007 12:00pm
Justin (mail):
Without getting us further afield, I tend to take Peter Strauss's (Columbia Law School)'s view of Chevron, and he's going to have said anything far clearer than I have, and I never meant to imply that once the agency's interpretation has been upheld it cannot then change its interpretation.
7.26.2007 12:13pm
martinned (mail) (www):

Another AEDPA issue I've always wondered about. Given the deferential standard of review it requires, how will the courts ever be able to formulate any new "clearly established case law"?
7.26.2007 12:22pm
John C (mail):

There are pros and cons to AEDPA, but I personally don't find the idea of deferential review like that to be "ridiculous." Appellate courts are used to deferential standards of review, whether of facts ("substantial evidence" supporting a verdict) or of law (whether an agency's interpretation of an ambiguous statute is "reasonable" under Chevron). Such judgment calls can be hard, but that's why we call them judges.

The problem is that, as many posters here have said, AEDPA is not "deferential review" of a sort courts are accustomed to making. For instance, the rationale for reviewing a trial court's finding of fact for "clear error" makes much sense - the trial court is closer to the facts, has a presumed familiarity with the actors of the case, etc. And so a reviewing court can resort to fairly dependable tools of reason, logic, judicial experience, common sense, etc. to determine if a obvious mistake has been made by the trial court, while at the same time respecting the fact that the trial court is closest to the action and presumably is privy to some vagaries of information that might go unseen upon review.

Not so with AEDPA. Is there any principled way to determine what is "clearly established precedent"? Is one Supreme Court case enough? Does it need to be "super-precedent"? What about a recent case that overturned precedent? How much of that precedent is left? What about long-standing precedent that has been slowly chipped away by new cases? What, in other words, are the tools a reviewing court must use to determine what is "clearly established"? From my point of view, logic, reason, experience, and even common sense have no obvious role to play in such a review. It is ripe for instrumentalism.

And that is before we even get to "unreasonable application," which in my view supports no obvious tools of interpretation. My point is here that there is so much wiggle room in the AEDPA "standard" that it is no standard at all, and that it is silly to think that it will lead to consistent and convincing case law. The Getsy case is an example of that.
7.26.2007 1:01pm

The U.S. Supreme Court establishes it when it hands down a new decision
7.26.2007 1:02pm
Eric Muller (www):
7.26.2007 1:05pm
martinned (mail) (www):

So what would then happen is that the Supreme Court decides what the rule should be for the future, but refuses to apply it to the case at bar because it wasn't "clearly established" before? Wouldn't that make it an obiter dictum?
7.26.2007 1:07pm
John C,

I believe these are the answers to your questions:

Is there any principled way to determine what is "clearly established precedent"?

Yes, you look to what the Supreme Court has said the law is at the time of the state court case.

Is one Supreme Court case enough?

Of course.

Does it need to be "super-precedent"?

Of course not; there is no such category.

What about a recent case that overturned precedent?

The question is whether the law was followed at the time, not the new law.

How much of that precedent is left?

Again, the question is whether the law was followed at the time, not the new law.

What about long-standing precedent that has been slowly chipped away by new cases?

It doesn't matter what new cases have done: The issue is the clearly established law at the time.

What, in other words, are the tools a reviewing court must use to determine what is "clearly established"?

Look at what the Supreme Court said the law was at the time of the state court case.
7.26.2007 1:08pm
Doug Sundseth (mail):
Kerr's "The Constitution clearly allows a defendant to be retried and convicted after a first jury hangs" is a good example of abuse of a transitive verb.

Sorry; that's an incorrection*. "Hang" is also a fine intransitive verb, and "8 : to be uncertain or in suspense" is directly on point here. (Quoted definition from Merriam Webster's Collegiate Dictionary, 10th Ed.)

* For a discussion of the concept of the "incorrection", see this Language Log post.
7.26.2007 1:22pm
John C (mail):
Perhaps I worded my comment inartfully; what is the AEDPA solution if the Supreme Court precedent at the time of the state case does not squarely say "what the law is" at the time of the state court case? It can't be the case that there will always be crystal-clear SCUS precedent at the time of the state court case. And if there is not, how is a court to decide the correct precedent to apply? (the issue here is not when there is no SCUS precedent, but when that precedent is muddled).

Take, for example, the knock and announce rule. Prior to 1995, the SCUS had noted that the rule was "embedded in Anglo-American law" and invalidated evidence seized following a no-announce entry into a home. (Miller v. U.S.). But the Court had never had squarely held that knock and announce was an "element of the reasonableness inquiry of the Fourth Amendment" until Wilson v. Arkansas in 1995. So in the years between Miller and Wilson, was there clearly established precedent that a no-announce entry violated the Fourth Amendment? I think a strong argument could be made for both sides of that question - the Court admits in Wilson that it had never squarely addressed the question, but the Court in Miller threw out evidence gleaned as a result of the violation (a pretty clear indication to me). I think attorneys on both sides of that question would have equally valid arguments that each case was the first to establish the clear precedent. And what, at any given point, were the "clearly established" constitutionally-based exceptions to the knock and announce rule, if the very grounding of the rule in the Fourth Amendment was questionable in the time between the cases?

And that doesn't even begin to address what would be an "unreasonable application" of that precedent - difficult enough to determine if there is (objectively) clear precedent, totally impossible (I would submit) if there is not. What can be an unreasonable application of a rule who's contours are uncertain? You might answer by saying that if there can be no unreasonable application if there is no clear precedent (and thus no habeas relief for the petitioner), but I think that a judge could easily say (in a situation like above) that each case was first to hold clearly on the issue. And who's to say otherwise? You say that judges must judge, that that is there job, but I'd like there to be a consistent and convincing framework for making those decisions, and I don't think AEDPA provides it.
7.26.2007 2:10pm
John C,

Maybe I'm missing something, but again, I don't think these issues are so hard. "Clearly established" means, well, "clearly established." It not only needs to be established, it needs to be so clearly established that to ignore it in these facts is an unreasonable application of the clearly established law. In your example, the law is really unclear; there was a hint that maybe the Supreme Court might explicitly recognize a point someday, but it hadn't yet. Thus the law was not clearly established. (Maybe the 4th isn't the best example because there's no suppression remedy and these claims aren't recognizable in habeas anyway, but that's my take on the conceptual point.)
7.26.2007 2:29pm
Oren (mail):

Suppose Alice and Bob break into Charlie's house to rob him and one of them ends up shooting Charlie. Can the state try both for murder and claim, at both trials, that the defendant pulled the trigger?
7.26.2007 4:05pm
Doug Sundseth (mail):
Before the jury reports, the verdict is uncertain. But when the report is delivered the probability waveform collapses. In the case of a hung jury, the waveform collapses to "no verdict" because the jury is uncertain. (This could be described as Schroedinger's jury.) Thus, the vi form is entirely appropriate.

Also, the vi sense of "hang" with reference to juries is well established in edited prose; examples are easy to find. I assume from your comments that this usage would be a violation of your personal stylebook, but your stylebook preference need not be construed as controlling, and, in fact, I do not so construe it.
7.26.2007 4:20pm
Doug Sundseth (mail):
"Are you then willing to explain why we say "hung jury" rather than "hanging jury"?"


"I realize consistency has no place in the law, but in English usage, a "hung" jury is one that is "hung," making the verb in question a transitive one."

Your usage here is transitive. There is also an intransitive usage. Both are used in edited prose (heck, probably poetry, too, though I don't have a cite).

"Check out some 7th-grade grammar."

I have a snarky response, too. In accordance with the rules at the bottom of the page, though, I'll keep it to myself.
7.26.2007 6:17pm
Doug Sundseth (mail):
Please pardon my error. Your usage here is not intransitive, since you aren't using any form of "hang" as a verb. In "hung jury", "hung" is an adjective. In "[jury] that is hung", "hung" is a direct object of "is". Neither of these directly addresses the issue of whether the apposite use of "to hang" should be vi or vt.

For an actual vt usage, try, "The evidence was ambiguous enough that it hung the jury." Note that the past tense is appropriate to the tense of the sentence. If you were in the room, you might be able to say, "The evidence is so ambiguous that it is hanging the jury."

This is a vi form: "The evidence is so ambiguous that the jury is likely to hang." This does present a possible confusion, but that's common in English. In any case, the vt form is entirely grammatical.
7.26.2007 6:35pm
Doug Sundseth (mail):
Geez, late in the day and responding too fast, so I keep making basic errors.

In "[jury] that is hung", "hung" is an adjective again.
7.26.2007 6:39pm
Zacharias (mail):
Jimbino is right, Doug.

In addition to your confusion concerning the transitive/intransitive distinction, you seem to be confused about present and past participles of verbs.
7.26.2007 7:37pm