MELVIN WOFFORD, Petitioner-Appellee, v. JEFFREY WOODS, Warden, Respondent-Appellant.
No. 18-2367
UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT
Decided and Filed: August 13, 2020
RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b). File Name: 20a0259p.06. Argued: October 17, 2019.
Appeal from the United States District Court for the Eastern District of Michigan at Detroit. No. 2:16-cv-13083—Laurie J. Michelson, District Judge.
Before: BOGGS, BATCHELDER, and DONALD, Circuit Judges.
COUNSEL
ARGUED: David Porter, OFFICE OF THE MICHIGAN ATTORNEY GENERAL, Lansing, Michigan, for Appellant. Colleen P. Fitzharris, FEDERAL COMMUNITY DEFENDER, Detroit, Michigan, for Appellee. ON BRIEF: David Porter, OFFICE OF THE MICHIGAN ATTORNEY GENERAL, Lansing, Michigan, for Appellant. Colleen P. Fitzharris, FEDERAL COMMUNITY DEFENDER, Detroit, Michigan, for Appellee.
BOGGS, Circuit Judge. The right to “trial by an impartial jury” guaranteed by the
Wofford was found guilty of murder in a Michigan court following the removal and replacement of a juror. While that juror was holding out against conviction at the time, the judge removed her because of her misconduct: she had violated his instructions not to discuss the case with anyone other than her fellow jurors by hiring a lawyer to address the court about
In assessing this appeal, we first review the factual determinations that the federal district court made in reviewing the factual findings of the MCOA. This review shows that, while the juror was a holdout, she was not removed for this reason, but instead because of her misconduct.
Then, we turn to review whether the district court was legally correct in granting a writ of habeas corpus on these facts. The recent Supreme Court decision in Ramos provides guidance on the
I. Factual and Procedural History
A. The Murder
In June 1993, Thomas Gilmore was strangled to death outside of his employer‘s offices late at night. Gilmore, a surveyor, worked for a company located in a building that also contained a roofing company. Police found evidence suggesting that a person or persons familiar with the interior layout of the roofing company had broken into it, stolen a circular saw, and attempted to force open an inner office where more valuable equipment and cash were kept. The point of entry appeared to be a previously broken window that was covered with sheet metal held up by duct tape and screws. Police found two hairs stuck to the duct tape that held the sheet metal in place and, on a wall and work bench, two droplets of blood that they suspected came from the burglar after he cut himself on some of the broken glass while entering.
The police theorized, and prosecutors would later argue, that Gilmore had an encounter with the person(s) involved, either while taking a break outside from his late-night work (as he sometimes would) or after he was lured outside. Gilmore had been beaten severely and then strangled with a rag or piece of clothing. Gilmore‘s assailant used the handle of a hammer, taken from the roofing company, either in Gilmore‘s beating or to assist in his strangulation,
B. The Trial
Wofford was tried for murder in a Michigan state court in August 2013. The prosecution not only relied on the physical evidence and deductions outlined above but also introduced evidence that Wofford had needed money for a drug habit and that the roofers’ pay had recently been delayed by the company. A former coworker testified that Wofford had been out drinking with him on the night of the murder but mysteriously disappeared at about 3 A.M., only to reappear the next morning sleeping in the coworker‘s living room. Moreover, witnesses said that Wofford did not return to work the day after the murder or the day following, and may have never returned again. As some of the evidence at the scene of the crime did not point to Wofford, the prosecution offered a theory in the alternative that he had had an accomplice.
The defense relied primarily on raising points of reasonable doubt. The initial investigation of the burglary had been sloppy, as it was at that time a low-priority crime; Gilmore‘s body was not discovered, hidden in the bushes outside, for another 24 hours after the break-in was reported. By the time the body was discovered, the crime scene was contaminated. The defense argued that it was natural for some of Wofford‘s blood to be in a place where he worked with sharp tools and construction equipment, and that he could easily have shed some of his “long blond hairs” while working, which subsequently were carried by the air and got stuck in tape. The prosecution countered with testimony indicating that the blood droplets were fresh and invited the jury to observe that nothing else, not even dust, was attached to the duct tape, which implied that it had been freshly torn from the metal that covered the window when the hairs stuck to it. The defense further argued that only a used circular saw worth $30 had been stolen, and the burglary would thus have been a misdemeanor; therefore, concealing his identity would be a poor motive for the burglar to commit murder. Because Gilmore had been so badly beaten, the defense inferred that he must have fought back. Yet Wofford‘s coworker testified that he was in the same clothes the next morning as the night before, which did not appear bloody or torn, and witnesses who saw Wofford in the days after the murder did not recall seeing bruises on his hands or face. Nor was his blood identified among the samples taken from Gilmore‘s body. Finally, the defense pointed to two sets of alternative suspects: another tenant of the building with a history of violence, to whose son Gilmore owed money, and two men who had been convicted of a burglary and murder in a nearby town at around the same time.
When the judge swore the jury at the beginning of the case, he instructed them that:
You must not discuss the case with anyone, including your family or friends. You must not even discuss it with the other jurors until the time comes for you to decide the case. When it is time for you to decide the case I will send you to the jury room for that purpose. Then you should discuss the case among yourselves but only in the jury room and only when all the jurors are there. When the trial is over you may, if you wish, discuss this case with anyone.
On Monday, August 26, after the jury had been deliberating for about seven total hours, the jury foreperson sent a note to the judge, reading, “We are eleven to one with no chance of the one moving their view.” The judge consulted with the lawyers for both sides, gave the jury a deadlock instruction, and sent them back to continue deliberating. Just an hour after he sent them back in, his clerk received a note from another juror (not the foreperson): “Excuse me, judge, one of our jury‘s [sic] doubts are unreasonable, what do we do?” The judge did not respond directly. Instead, realizing that the jury may not have been given a standard instruction, he summoned the jury and read them the following:
If you want to communicate with me while you‘re in the jury room please have your foreperson write a note and give it to the court clerk. It is not proper for you to talk directly with the Judge, lawyers, court officers or other people involved in the case. As you discuss the case you must not let anyone, even me, know how your voting stands. Therefore, until you return with a unanimous verdict do not reveal this to anyone outside the jury room.
Neither of these notes revealed the gender of the juror referred to. Nothing further of note happened that day.
The next morning (Tuesday) was quiet. Then, in the mid-afternoon, the jury sent three notes to the judge within the space of half an hour. The first read:
We the jury have a member who is not cooperating and refuses to deliberate or prove to us her vote. She just wants a hung jury. She also stated she had looked up the phrase to see what it meant before deliberation even started. Please advise us on what to do in this case. Thank you!
The second said:
We Have a Jury member who SERIOUSLY doesn‘t understand what Reasonable Doubt is!! We have a hung jury and we need instructions!!! Help!!!
The third note, meanwhile, asked a question about the law of aiding and abetting.2 At this point, the court heard from both counsel. The prosecution suggested that the juror referenced in the second note
At this point, the defense attorney made a mistake. Operating without a transcript, he mistakenly ‘recalled’ that one of the previous day‘s notes had referred to the holdout juror by a masculine pronoun (it had not, speaking only of “one of our jury‘s [sic] doubts“). Thus, he concluded that the use of “she” in the first note that day meant that the composition of the jury‘s
vote had changed since the original note. Therefore, he equivocated as to whether the court should declare a mistrial or reread a deadlock instruction. The misperception thus created would continue to plague the proceedings.
The judge decided to give a series of instructions to the jury: an aiding-and-abetting instruction, a reasonable-doubt instruction, an instruction not to seek outside information and to follow the law, and a second deadlock instruction. The defense made no objection. Thus ended the third full day of deliberations. On Wednesday—on the afternoon of which the judge was unavailable—nothing more came out of the jury room (aside from a request to see an item of evidence).
Just before noon on Thursday, an attorney presented himself in court and said that he had been hired by one of the jurors, Juror M. He explained that his client had contacted him the evening before:
She said that she needed my help and that she was being harassed and verbally abused on a jury that she was sitting on. And we didn‘t discuss any of the facts of the case or discuss anything about a vote or anything like that, I just advised her that she should notify the Court through a note to the bailiff . . . . And then, you know, later on I talked to her . . . I said look, if you‘re uncomfortable . . . it was obvious she was uncomfortable . . . I said, you give me permission, I can notify the Court, and after several discussions last night she instructed me to notify the Court.
. . . .
I didn‘t know any of the facts. I don‘t—other than what was advised to me while we were in the hallway by Mr. Lynch [defense counsel], so I know nothing about the case other than what Mr. Lynch had told me, and of course, none of that had—had been or has been discussed with [Juror M].
The court asked both counsel why he should not dismiss the juror for cause, as contacting the lawyer was “a flagrant violation of the Court‘s instructions, cautionary instructions, which were not to discuss the case at all with anyone outside of the jury room . . . .” “Let‘s say she said and it‘s eleven to one or gone in detail, I think that would be certainly cause to excuse her from the jury without a blink of an eye.” Even assuming that Juror M had said nothing about the substance of the case, her discussion of “the climate of the jury room” may well “constitute[] cause to excuse her.”
The defense at this point requested that the judge declare a mistrial on the grounds that replacing the juror would mean “there‘s going to be issues whether it‘s a coercive verdict.”3 The prosecutor argued
[I]f I agree . . . that we don‘t know what the jury is doing, but also agree that there‘s been a violation -- juror misconduct in contacting a lawyer, why isn‘t the proper remedy to excuse her for cause? I have two alternates that are out there. Why -- why not bring -- if somebody got hit by a bus this morning I would bring in the alternate juror. I wouldn‘t -- I don‘t know what the difference is. Why should I declare a mistrial?
. . . [W]e don‘t even know that the person that has written the [note] is the holdout, we don‘t even know there is a holdout. For all we know it‘s six/six. . . . Or eight/two or eight/four or whatever it is. We don‘t know anything except for the very first note what the -- what it was. And as we noted before, the gender‘s changed. The first note was one gender, the second note was another gender. So for all we -- we have -- this is just pure rank speculation at this point. So why isn‘t the . . . so my my question is, the Court‘s thinking is, I‘ve got a juror that‘s violated the instructions, she -- obviously she felt compelled to do it. She may -- who knows what she‘ll feel compelled to do with-- in the future if I keep her on the jury. Why don‘t I strike her for cause, bring in the alternate?
Defense counsel conceded, “I agree that the pronouns change . . . .” At the same time, he argued, one person seemed to have become the target of the last two notes. It was “fair to say” the atmosphere in the jury room had become one of “frustration” and “animosity[.]” Thus the defense maintained that a mistrial should be declared, because otherwise the replacement juror would feel coerced.
The judge consulted
I reject the speculation that -- that there is sufficient cause at this point to warrant a mistrial. That is based on a speculation and conjecture about what‘s going on in the jury room. What we do know is that there have been some notes, and we‘ve discussed those ad nauseam, that one juror has retained counsel and has -- that counsel has made a statement on the Record. I find that that is a flagrant violation of this Court‘s jury instructions, that that constitutes good cause to excuse that juror for cause, and I so by order that she be excused. How that affects the complexion of the jury deliberations is unknown and is not the province of any individual to pierce into.
Once the alternate was seated, the judge ordered the jury to “begin your deliberations anew.” The reconstituted jury returned a guilty verdict within one and a half hours. Wofford received a mandatory sentence of life without parole.
C. The Appeal
Wofford appealed to the Michigan Court of Appeals, where he raised two issues: a challenge to the removal of Juror M, which he framed as a “den[ial]” of “his Sixth Amendment right to have his jury decide his fate” and a sufficiency-of-the-evidence
The case of United States v Thomas, 116 F.3d 606, 621 (CA 2, 1997), provides helpful guidance: “To remove a juror because he is unpersuaded by the Government‘s case is to deny the defendant his right to a unanimous verdict.” Also, “If the record raises any possibility that the juror‘s views on the merits of the case, rather than a purposeful intent to disregard the court‘s instructions, underlay the request that he be discharged, the juror must not be dismissed.” Id. at 622, n. 11 (internal citations omitted).
Wofford then discussed a defendant‘s right to have his case heard by the original jury selected, referring to several Michigan precedents,5 as well as two United States Supreme Court double-jeopardy cases.6 Wofford, his counsel concluded, had been “deprived of his right to have his case fairly tried and determined by his jury.”
The MCOA rejected this argument. It assessed Wofford‘s claims relating to the juror removal in light of People v. Tate, 624 N.W.2d 524, 529 (Mich. Ct. App. 2001), from which it quoted the following rule:
[W]hile a defendant has a fundamental interest in retaining the composition of the jury as originally chosen, he has an equally fundamental right to have a fair and impartial jury made up of persons able and willing to cooperate, a right that is protected by removing a juror unable or unwilling to cooperate. Removal of a juror under Michigan law is therefore at the discretion of the trial court, weighing a defendant‘s fundamental right to a fair and impartial jury with his right to retain the jury originally chosen to decide his fate.
People v. Wofford, No. 318642, 2015 WL 1214463, at *2 (Mich. Ct. App. Mar. 17, 2015) (brackets in original) (quoting Tate, 624 N.W.2d at 529 (cleaned up)). Analyzing Wofford‘s case under this rubric, the MCOA made several factual findings. First, that “[c]ontrary to defendant‘s assertion on appeal, the trial court did not remove [Juror M] because she was a lone hold-out who was standing in the way of conviction.” 2015 WL 1214463 at *2. As the MCOA saw it:
Early on in the deliberations, the jury informed the court that they were 11 to 1. However, the jury did not reveal which way the majority was leaning or the identity of the hold-out. Several days of continued deliberation then ensued. As the trial court found, there was no evidence at the time of [Juror M‘s] removal what way the jury was leaning, or the potential division.
Ibid. (emphasis added). Second, the MCOA found that: “[T]he record clearly demonstrates that the trial court removed this juror because she flagrantly violated the court‘s instructions by discussing the deliberations with a non-juror.” Ibid. Observing that “neither party contests”
Juror M‘s “blatant disregard for the court‘s instructions, nor that it implicated her ability to
Finally, the MCOA observed that Wofford had argued that “the trial court still should have granted a mistrial because of the danger that the alternate juror would be influenced when faced with 11 jurors who were convinced of his guilt.” Id. at *3. To this, it responded:
Yet . . . there is no evidence regarding what way the jury was leaning at the time of [Juror M‘s] removal. Thus, to conclude that there was only one hold-out, or that it was [Juror M], is mere speculation. Significantly, the trial court properly instructed the newly configured jury that it must deliberate anew, and disregard its prior deliberations. See
MCR 6.411 ; Tate, 244 Mich.App at 567 (“the jury in such situations should be instructed to begin deliberations anew.“). . . . “Because it is . . . pure speculation that the alternate juror was coerced, we conclude that defendant was not prejudiced by the substitution of the alternate juror.” People v. Mahone, 294 Mich.App 208, 218; 816 NW2d 436 (2011).
Id. at *3 (third ellipsis in original). The MCOA thus held that Wofford‘s rights had not been violated, and the Supreme Court of Michigan denied him leave to appeal. People v. Wofford, 869 N.W.2d 598 (Mich. 2015).
D. The Habeas Petition
Wofford then filed a petition for a writ of habeas corpus in the United States District Court for the Eastern District of Michigan. The district court applied AEDPA deference when reviewing three factual findings: that there was “no evidence” of “what way the jury was leaning” when Juror M was removed; that the trial court did not remove Juror M because of her views; and that the judge removed Juror M for “flagrantly violating the court‘s instructions.” The district court found the first to be rebutted by clear and convincing evidence: Once the confusion over the gender in the notes is removed, it is clear that “all along the way, there was but one holdout,” namely Juror M. The court said that the post-removal guilty verdict confirmed this, especially as it was obtained within ninety minutes. Despite expressing “skeptic[ism,]” the court held that the other two factual findings could not be rebutted by clear and convincing evidence. Then the district court turned to the legal analysis. It held that AEDPA‘s deferential standard of review did not apply to the MCOA‘s legal analysis because that court had “never addressed ‘the merits’ of Wofford‘s claim that his Sixth Amendment right to a unanimous jury was violated.” It evaluated whether the MCOA had addressed the merits of Wofford‘s
Freed from AEDPA deference, the district court proceeded to apply the test from Symington, finding that the record disclosed a reasonable possibility that the impetus for Juror M‘s dismissal stemmed from the merits of the case and that “Wofford was thus deprived of his Sixth Amendment right to a unanimous verdict.” The writ of habeas corpus was granted, and the state was ordered to re-try Wofford or release him within 90 days.11
This appeal timely followed—which would normally be the end of the factual narrative. But something unusual happened two months later. In January 2019, the district court issued a further order addressing the state‘s motion for a stay pending appeal and Wofford‘s motion to be released on bond. It granted the former and denied the latter. In doing so, the court substantially
revised the basis for its initial opinion. The state had argued in its brief on the motion that the district court‘s initial ruling was in error, as the
The Court‘s holding rests on the test of Symington. . . . And that very same test is used to decide whether a criminal defendant‘s right “against improper interference with jury deliberations” has been violated. See Williams v. Cavazos, 646 F.3d 626, 646 & n.17 (9th Cir. 2011), rev‘d on other grounds sub nom. Johnson v. Williams, 568 U.S. 289 (2013). And that right has been incorporated against the states[.]
For this proposition, the district court then quoted the following excerpt from Williams v. Cavazos:
The
Sixth Amendment protection required by Brown and foreshadowed by Symington is against improper interference with jury deliberations. Freedom to deliberate without coercion is a necessary component of “the interposition between the accused and his accuser of the common-sense judgment of a group of laymen,” which is “the essential feature of a jury.” Williams, 399 U.S. at 100, 90 S. Ct. 1893. This component of theSixth Amendment is thus incorporated with respect to the states through the Due Process Clause of theFourteenth Amendment . In Weaver v. Thompson, 197 F.3d 359 (9th Cir. 1999), for example, we held that improper coercion had broken a jury deadlock to produce the ten-to-two vote needed to convict under Oregon law, and granted habeas as a result. Similarly, the deadlock-breaking dismissal of a holdout juror on an improper basis is an unconstitutional form of interference with deliberations.
Williams, 646 F.3d at 647 n.17. Therefore, the court concluded that “while the Warden might be correct that the Court misstated the
This order, in attempting to fix one problem, created two others. The first is obvious: the district court, having come very near to acknowledging a fundamental flaw in the earlier ruling, attempted to recast it not through a rehearing or reconsideration, but rather through a separate order. The second is that, as we shall see, the Ninth Circuit opinion in Williams is highly doubtful law. It was not “rev[erse]d on other grounds,” but rather on the exact same grounds as in this case.
II. Standard of Review
Appellate review of a district court‘s decision not to apply AEDPA to a petition for a writ of habeas corpus involves the application of multiple standards nested within one another like Russian dolls. We review the district court‘s decisions of law de novo and of fact for clear error. Wheeler v. Simpson, 852 F.3d 509, 514 (6th Cir. 2017); Scott v. Houk, 760 F.3d 497, 503 (6th Cir. 2014). But federal courts in turn have a statutorily prescribed standard by which it is to assess the decisions of the state courts. Federal courts are not to grant a writ of habeas corpus to state prisoners “with respect to any claim that was adjudicated on the merits in State court proceedings” unless that state court ruling:
- resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or
- resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.
Thus, the hurdle for a district court to overturn the rulings of a state court is very high. To be sure, “[t]he standard is demanding but not insatiable,” Miller-El v. Dretke, 545 U.S. 231, 240 (2005), and “[d]eference
Thus, to put it all together, we examine de novo whether the district court correctly ruled, on the law, as to whether the state court arrived at a conclusion on federal law opposite to that reached by the Supreme Court or unreasonably reached a different conclusion from the Supreme Court when confronted by a materially indistinguishable fact pattern. And we review for clear error the district court‘s finding of facts, but the defendant must have shown the district court, by clear and convincing evidence, that the state court‘s determination of facts was unreasonable.
All that being said, AEDPA‘s demanding system of deference does not apply to the district court in one circumstance: if the state court has, through “sheer inadvertence” overlooked a plaintiff‘s federal claim. Johnson, 568 U.S. at 302-03. This must be shown “very clearly” by the evidence, id. at 303, because usually we presume that a state court—whether it has addressed only some of the issues presented on appeal or even none at all—has adjudicated all claims brought on the merits. See id. at 298-303; Richter, 562 U.S. at 98-100. In the rare circumstance where the state court has inadvertently overlooked a claim, the district court can review that claim de novo, i.e., freed from the constraints of AEDPA deference. Johnson, 568 U.S. at 303. We would then apply the usual standards of appellate review—de novo on the law and clear error on the facts—to the district court‘s rulings. Scott, 760 F.3d at 504 (citation omitted).
III. Factual Analysis Under AEDPA
We turn first to review the district court‘s determinations of fact under AEDPA; once we have established what the facts are, we will proceed to the law. The district court reviewed three factual determinations made by the Michigan courts for whether the defendant had shown them to be wrong by clear and convincing evidence. They were: (1) that, according to the MCOA, there was “‘no evidence’ of ‘what way the jury was leaning’ or ‘the potential division’ at the time Juror M was removed” (quoting Wofford, 2015 WL 1214463, at *2); (2) that “the trial court did not remove [Juror M] because she was a lone hold-out who was standing in the way of conviction” (quoting Wofford, 2015 WL 1214463, at *2); and (3) that “the trial court removed this juror because she flagrantly violated the court‘s instructions.
A. The Record Shows There Was One Holdout All Along
The secrecy of deliberations is one of the most vital components of the jury system, and courts strive mightily to maintain it. But just because we are forbidden to look into the black box does not mean we should deliberately ignore signals coming out of it. The MCOA, whose decision—rather than that of the trial court—was reviewed by the district court in this case, repeatedly asserted that there was “no evidence” of “what way the jury was leaning” or the “potential division” when Juror M was removed. Wofford, 2015 WL 1214463 at *2-3.
As we have seen, there was persistent confusion in the courtroom over whether one juror or two had held out, as a result of the defense attorney‘s misremembering the first note as
containing a male pronoun. This goes to the subjective understanding of the judge at the time he made his removal decision, which matters for other reasons. See infra, p. 20. But the MCOA was speaking to what is now shown by the evidence, not what was then in the judge‘s mind. And the evidence shows a clear picture.
Fairly early in the deliberations—at most, 7 hours in—a deadlock set in. The jury was split 11-1 “with no chance of the one moving their [sic] view.” The next day, two notes mentioned a hung jury. It seems very unlikely that the jurors would have continued to consider the situation a deadlock had the composition of the vote switched from Monday to Tuesday. Furthermore, as the federal district court observes, both of the notes reference a single juror, and one of them makes it clear that this was a female juror. Two days later, the situation had become so acrimonious that Juror M hired an attorney to speak with the judge—in the process, revealing her identity and thus her gender, female. It seems improbable that this amount of pressure built up if the holdout juror had not been the same person the whole time.
Were we to imagine otherwise, that would mean at least ten people had changed their minds at least once. Putting aside the matter of comparative statistical likelihood, surely this would have shown a room full of jurors who were reconsidering their previously-held positions and changing their views—which seems entirely inconsistent with the atmosphere articulated in the notes, that of each side digging in and frustration mounting. Finally, of course, there is the fact that once Juror M was removed, the jury returned a guilty verdict within an hour and a half.13 Together, these facts demonstrate to us “clearly and convincingly” that, as the district court
As we have noted, judges are enjoined strongly to respect the secrecy of jury deliberations. What seems to have happened here was that both at the trial and appellate level, over-scrupulousness as to this commandment led perhaps to obliviousness as to what was going on. At trial level, this was compounded by, but also in large part excused by, the confusion over the gendered pronoun. On appeal, looking at the evidence without such confusion, the conclusion should have been plain. We therefore uphold the district court‘s finding the that the factual determination of the MCOA on this point was rebutted by “clear and convincing evidence.”
B. The Trial Court‘s Reasons for Removing Juror M
The district court expressed skepticism of, but ultimately let stand, the MCOA‘s determinations that “the trial court did not remove [Juror M] because she was a lone hold-out who was standing in the way of conviction” and that “the record clearly demonstrates that the trial court removed this juror because she flagrantly violated the court‘s instructions.” Wofford, 2015 WL 1214463, at *2. It examined both together, which makes sense, as each is the flip side of the other. To work through it logically, though, we should look at the latter determination first. If the trial court did not have at least a good independent reason to remove Juror M, the entire analysis would become different—all that would be left would be Juror M‘s holdout status, which would be an impermissible basis for removal. If the trial court did have a good, independent reason to remove Juror M, on the other hand, the inquiry then becomes whether this was pretextual or the actual reason for removal.
First, the trial court did have a good independent reason to remove Juror M. As we have seen, at the beginning of the trial, at the close of every day, and the end of the trial, the judge warned jurors not to talk to anyone about the trial. Juror M contacted an attorney about the deliberations in the case. The trial judge found “that that” was “a flagrant violation of this Court‘s jury instructions.”14 (Emphasis added.) Wofford tries to make hay out of the fact that the jury
instruction given on the Monday said only not to “talk directly with the Judge, lawyers, court officers or other people involved in the case.” But even if we read “lawyers” as “lawyers . . . involved in the case,” the omission of outside lawyers here is hardly surprising—as one of the defense counsel noted on the day the juror‘s attorney showed up, he had “[n]ever had this happen in 21 years.” And the omission of a catch-all phrase such as “and no one else” makes sense given the context of that instruction, which arose after a juror handed a note to the judge‘s clerk. It also should have been read in the light of the judge‘s constant daily reminders
The district court pronounced itself “skeptical” of the MCOA‘s findings of fact on this front because:
[I]t was obvious to all that from the time of the very first note, there was a single holdout. Consider too that the judge said, “I‘m not sure one way or the other at this point whether or not [Juror M‘s] violation constitutes cause to excuse her,” then, without any new insight from counsel on that particular issue, found that Juror M‘s conversation with an attorney “flagrant[ly]” violated his instructions. (R. 9, PageID.2128.) And if Juror M flagrantly violated the court‘s instructions by contacting an attorney, why were the notes from the jury indicating where they stood not also a “flagrant” violation of his instructions? Yet the foreperson was not removed.
Again, it is within the sound discretion of the trial judge to weigh questions such as which violations are severe enough that dismissal is required, as opposed to which can be handled by
other, less drastic means. It should also be self-evident that a judge can change his mind by taking some time to think about things, as well as—or instead of—listening to counsel. As to the first objection—that it was obvious from the start that there was a single holdout—the district court itself answers this a paragraph later: “[T]he trial judge‘s statements that a mistrial would be based on ‘speculation and conjecture about what‘s going on in the jury room’ and that Juror M had flagrantly violated his instructions suggest that, in the judge‘s mind, Juror M was not the holdout and that, in his mind, the problem with Juror M was that she violated his instructions.”
That brings us in turn to whether the court‘s finding of, or reliance on, the violation of the court‘s instructions was pretextual. The record reveals nothing to indicate that the judge bore an animus against Juror M, or that he was trying to influence the verdict, or that he was in any way behaving other than as a judge trying to apply the law and, as it requires, not rush a mistrial after a long and important criminal case. The mistake about the gender of the juror, while unfortunate, made it subjectively less likely (as the district court noted) that the judge was targeting Juror M in particular for her views. To be sure, we are—as was the district court—working from a cold record. But there is nothing here, beyond the bare facts—that there was a holdout, the notes from the jury, the appearance of counsel, and the judge‘s decision to remove that juror—to suggest that his removal decision was actuated by the juror‘s views on the case.
Moreover, in order to disturb the MCOA‘s findings on this front, we would have to find both that the district court was clearly in error in upholding them and that the defendant can rebut the MCOA by clear and convincing evidence. There is certainly not enough evidence to do so. Thus, the findings stand: we hold that the
IV. Legal Analysis
We begin by establishing the existence and limits of the constitutional right that underlies Wofford‘s claim, before turning to the doctrines that Michigan and federal courts have crafted to
protect that right. Then we examine how AEDPA affects our review of how the MCOA and the district court applied these doctrines.
A. Juror Removal and the Role of the Trial Judge
Since oral argument in this case, the Supreme Court has issued a decision that brings much-needed clarity to where the right in question before us is found in the Constitution. In Ramos v. Louisiana, 590 U.S. ___, 140 S. Ct. 1390 (2020), the Court confronted the question of whether the
Though the two are conceptually related, the constitutional right at issue in this case is not precisely the same as a unanimous-jury right. Wofford illustrates this well by arguing that if the state required a 10–2 vote (albeit this is now forbidden) and the jury were split 9–3, a judge could still not impermissibly remove and replace one of the three dissenting jurors so as to secure a guilty verdict. Thus, Ramos does not exactly resolve the issue before us. But it does provide significant guidance. The constitutional question at the heart of our case is whether and under what circumstances a judge can remove a juror either for the juror‘s substantive views or for misconduct or in a situation where there is both possible substantive disagreement and misconduct. Insofar as there are constitutional protections against certain impermissible juror-removal decisions (and, as we shall see, there are), they are part of the “requirements” that the Ramos Court recognized as inherent in the right to “trial by an impartial jury” secured by the
An examination of sources similar to those relied on by Ramos—in some instances, the same sources—illustrates the content and limits of this right. This not only allows us to identify the right asserted in this case with greater precision than was possible in the precedents that predate Ramos, but also shows the historical and logical reasons that
The jury trial had a long, convoluted development during the early Middle Ages, but it does not appear at any time that judges would routinely remove jurors due to their decisions.15 In 1367, it was established that they could not do so. At an assize in that year, a jury split 11–1, and the judges trying the case “took a verdict from these and imprisoned the twelfth.” Thayer, 88. As Thayer explains:
On moving for judgment, when counsel urged that it had formerly been adjudged in trespass that a verdict of eleven might be good, “and this we will show you by record,” Thorpe, C.J., said: “It is fundamental (la ley fuit fondue) that every inquest shall be by twelve ... and no fewer.... Though you bring us a dozen records, it shall not help you at all; those who gave judgment on such a verdict were greatly blamed.” Moubrey, J.: “As the verdict was by eleven and judgment cannot be rendered, sue out a new inquest and let the man imprisoned be discharged.”
Ibid. (ellipses in original) (citing Anonymous Case, 41 Lib. Assisarum 11 (1367), and 2 MATTHEW HALE, PLEAS OF THE CROWN 297 (1736)); see also Forsyth, 241. The Ramos Court cites this case from 1367 for the development of the principle of unanimity. See 140 S. Ct. at 1395 & n.11. But for our purposes, what matters is that the 1367 case held that a judge could not remove a juror due to a disagreement on the merits of the case. As a modern history confirms, verdicts were taken from eleven jurors before 1367, with six instances of jurors being imprisoned for being the holdout juror before that year, but thereafter the practice was regarded as illegal. David J. Seipp, Jurors, Evidences and the Tempest of 1499, in THE DEAREST BIRTH RIGHT OF THE PEOPLE OF ENGLAND: THE JURY IN THE HISTORY OF THE COMMON LAW 88 (John W. Cairns & Grant McLeod eds., 2002). See also Thayer, 88–89 n.4.
For several centuries, judges retained other ways of punishing those jurors who were thought to have given verdicts contrary to the evidence and/or the law. Primary among these was the fining or imprisonment of jurors for their verdict, until this practice was abolished in Bushel‘s Case, 124 Eng. Rep. 1006 (C.P. 1670). That is, while after 1367 the juror could not be removed from the jury, until 1670 he could still be imprisoned or fined after the fact. Courts also had tools that were not punishments as such but would nowadays be regarded as coercive. It was accepted practice that judges would cut off the jurors from food and drink until they returned a unanimous verdict. Thayer, 155 & n.2; 2 Hale, 297. Judges could also cart the jury around from town to town as the judges rode circuit, until the jury came to a decision. 2 Hale, 297; Forsyth, 241. Such methods usually produced unanimity of result, if not of conscience.16 American courts
Judges instead came to rely, starting in the seventeenth century, on their authority to declare a mistrial as a way of dealing with jury misconduct. See Hinton v. United States, 979 A.2d 663, 672 & nn.17 & 20 (D.C. 2009); Thayer, 169. But while this method, unlike a fine or a trip to jail, did not harm the jurors, it fell hard on someone else: the defendant. This was a particular concern in criminal trials, where there was a well-founded concern that certain judges would “withdraw[] a juror’ and declar[e] a mistrial in order to thwart an acquittal.” Hinton, 979 A.2d at 673. “In reaction to such abuses, common law judges eventually settled on a generally applicable rule of practice: no juror could be withdrawn, and no jury could be discharged, unless it was necessary to do so.” Ibid. (collecting sources from 1698 onward).
In England, assessing the “necessity” involved was left to the trial court‘s discretion—which still left potential for abuse. In America, things were different:
In early American jurisprudence, more plainly than in coeval British decisions, the rule against unnecessary mistrials was a right the defendant could assert. In part, that was because “state courts... blend[ed] the rule against needless discharges of juries into the guarantee against double jeopardy contained in the Federal and State Constitutions,” despite the contrary analysis of roughly contemporary English jurists, who traced the rule to the maxim that once constituted, a jury could not be discharged until it had returned a verdict.
Id. at 674 (footnotes and citations omitted). That said, the judge‘s authority to remove a juror for good reason was long established: Already, in Hale, we read that “[t]he justices at common law may upon a just cause remove a juror after he is sworn.” 2 Hale, 296. Among circumstances that constituted good cause were not only death, illness, family emergency, or the discovery that a juror was also a witness or grand juror,18 but also certain types of juror misconduct.19
One further source of law on juror removal needs to be mentioned. Redoing a trial after a juror had been removed had always been recognized as onerous and, as trials became more complex, it became a bigger burden. See Hinton, 979 A.2d at 675. (This was true notwithstanding the practice of retaining the eleven remaining jurors and selecting a new one from the general public. Ibid.) Thus, starting in the late nineteenth century, states—and, in 1932, the federal government—passed laws providing for alternate jurors. See id. at 675–76 & n.40. This in turn gave rise to court rules governing when and how the court could remove an original juror and substitute an alternate, particularly
Due to the nature of case-by-case jurisprudence, not all of the strands of doctrine implicated by the removal of a juror were developed at once. Thus, the question of how the declaration of a mistrial did or did not implicate the right against double jeopardy was developed quite early, with a landmark Supreme Court case in 1824.22 Ramos, on the other hand, was decided
B. Michigan Law on Juror Removal
As a result of the history we have just surveyed, it has not been clear where to ground the analysis of a trial court‘s decision to remove and replace a juror. There are several ways to look at the problem. It can be thought of as a legal question about double jeopardy: Are you trying the same person before two different juries? Or an appeals court can ask when or whether the judge should have declared a mistrial, which implicates double-jeopardy jurisprudence but also the inherent power of appellate courts to oversee the discretionary judgments of trial courts. But removal and replacement can also be regarded as matter of administering the federal or state rules of criminal procedure. Or it can be treated as a question of
In assessing Wofford‘s complaint, the Michigan Court of Appeals relied on People v. Tate, 624 N.W.2d 524. In Tate, the defendant argued that a juror‘s removal had to be shown to be for “just cause.” The MCOA ruled that this was not so; rather that the basis for such a removal rested in the discretion of the trial judge, “weighing a defendant‘s fundamental right to a fair and impartial jury with his right to retain the jury originally chosen to decide his fate.” Id. at 529. In reaching this conclusion, the Michigan court drew heavily on People v. Dry Land Marina, Inc., 437 N.W.2d 391, 394 (Mich. Ct. App. 1989). See Tate, 624 N.W.2d at 529. These cases share in the overall trend of being clear on the rules for jury removal but mixing the doctrines from which they are derived.
Dry Land begins by analyzing strictly the removal of the juror, before reaching the replacement of the juror. It does so with reference to double-jeopardy precedents. See 437 N.W.2d at 392 (citing Jorn, 400 U.S. 470, and Gardner, 194 N.W.2d 62). Mindful of the federal precedents on mistrial going back to Perez, the court held that Michigan courts must use “less drastic alternatives” to declaring a mistrial if possible. Id. at 393. One of these less drastic alternatives is the empanelment of an alternate juror, a practice the Dry Land court approved. Ibid. But at the time, the Michigan rule under which the trial judge had empaneled the alternate authorized such replacement only in instances where the jury had not yet begun deliberating, as did the parallel federal
In assessing this question, the Dry Land court relied on a series of federal cases in which federal courts confronted similar questions about whether admitted contraventions of
Dry Land and several of the federal cases just discussed do consider the danger that the power to remove a juror can be abused. They do so by distinguishing the Ninth Circuit case of United States v. Lamb, 529 F.2d 1153 (9th Cir. 1975) (en banc).25 The facts of Lamb seemed to show not merely a misapplication of
Dry Land, citing those cases, adopted the same rule.26 437 N.W.2d at 394. The
The language of F.R.Crim.P. 24(c) [and MCR 6.102(A)] is not constitutionally mandated. Therefore, there is no violation of a defendant‘s right to trial by a fair and impartial jury when an alternate juror is recalled and substituted for a deliberating juror excused by the trial court. United States v. Phillips, 664 F.2d 971, 992–993 (C.A. 5, 1981) cert. den. sub. [sic] nom. United States v. Meinster, 457 U.S. 1136, 102 S. Ct. 2965, 73 L.Ed.2d 1354 (1982).
Dry Land, 437 N.W.2d at 394 (brackets in original). Thus, after Dry Land, the Michigan jury doctrine combined both
C. Federal Law on Juror Removal (and a Word on Jury Nullification)
As the district court in this case noted, several federal courts have imposed a prophylactic rule to deal with removal situations like that before us. In 1987, the District of Columbia Circuit addressed an appeal from a federal criminal trial in which a juror had been removed. Brown, 823 F.2d at 592. The juror had told the judge, first via note and then in court, that he was unable to perform his duties because “I can‘t go along with the [RICO] act.” Id. at 594. But upon further questioning, he also stated that “[i]t‘s the way [the law is] written and the way the evidence has been presented.” Ibid. The juror was removed, and the judge directed the remaining eleven jurors to deliberate as a jury of eleven under
The court noted, however, that “the reasons underlying a request . . . will often be unclear.” Id. at 596. Such situations therefore pose a problem. Moreover, the panel reasoned, it would damage the secrecy of jury deliberations for the judge to make searching inquiries of the juror(s) in such a situation. Ibid. Therefore, it crafted a prophylactic rule: “If the record evidence discloses any possibility that the request to discharge stems from the juror‘s view of the sufficiency of the government‘s evidence, the court must deny the request.” Ibid. “Any other holding,” the court held, “would fail to protect adequately a defendant‘s right to be convicted only by a unanimous jury.” Ibid.
Approximately a decade later, the Second Circuit and the Ninth Circuit confronted similar situations to that in Brown. Thomas, 116 F.3d at 606; Symington, 195 F.3d at 1082. In each of these cases, a juror, who turned out to be the holdout juror, was removed during a federal criminal trial. Thomas, 116 F.3d at 612; Symington, 195 F.3d at 1084. Again, there was evidence that pointed toward legitimate reasons for dismissal (in Thomas, both an inability to deliberate and a desire to nullify; in Symington, problems deliberating that may have been age-related), but there was also evidence that the juror had doubts about the prosecution‘s case. Thomas, 116 F.3d at 609–12; Symington, 195 F.3d at 1083–84. The district court in Thomas had dismissed the juror under
The Third and Eleventh Circuits have adopted a variation of the Brown-Thomas-Symington rule. United States v. Kemp, 500 F.3d 257, 304 (3d Cir. 2007); United States v. Abbell, 271 F.3d 1286, 1302 (11th Cir. 2001). So too have at least three states. State v. Elmore, 123 P.3d 72, 77–78 (Wash. 2005); People v. Gallano, 821 N.E.2d 1214, 1223–24 (Ill. App. Ct. 2004); Garcia v. People, 997 P.2d 1, 7 (Colo. 2000).
Moreover, as the Brown-Thomas-Symington rule became settled, courts have refined it. Most notably, the D.C. Circuit has recently clarified that even under Brown, “if the court forms an independent, good-cause justification for removing the juror that bears no ‘causal link’ to the juror‘s ‘holdout status,’ the court may excuse the juror even if the juror ‘independently had doubts about the sufficiency of the evidence.‘” United States v. McGill, 815 F.3d 846, 869 (D.C. Cir. 2016) (citing United States v. Ginyard, 444 F.3d 648, 652 (D.C. Cir. 2006)). See also United States v. Edwards, 303 F.3d 606, 633 (5th Cir. 2002).
As this review of the federal cases has shown, there is a lurking specter in the background of this issue: jury nullification. The question has periodically arisen whether the jury had the power to construe the law as well as the facts. In nineteenth-century America, this view was embraced particularly by Jacksonian populists, and, notwithstanding its repudiation by the Supreme Court in Sparf and Hansen v. United States, 156 U.S. 51, 102 (1895), it has continued intermittently to be a subject of controversy up to our own time. See Thomas, 116 F.3d at 614–16 (collecting sources).
Nullification is particularly controversial in this country because it has played a role in America at its worst and its best. On the one hand, nullification has been associated with some of America‘s most important moments of civil disobedience. “The case of John Peter Zenger, the publisher of the New York Weekly Journal acquitted of criminal libel in 1735, and the nineteenth-century acquittals in prosecutions under the fugitive slave laws, are perhaps our country‘s most renowned examples of ‘benevolent’ nullification.” Thomas, 116 F.3d at 614; see also United States v. Dougherty, 473 F.2d 1113, 1130 (D.C. Cir. 1972). Moreover, in less-famous trials, the practice of nullification is thought to “introduce[] a slack into the enforcement of law, tempering its rigor by the mollifying influence of current ethical conventions.” Thomas, 116 F.3d at 615, quoting U.S. ex rel. McCann v. Adams, 126 F.2d 774, 776 (2d Cir. 1942) (Hand, J.), rev‘d on other grounds, 317 U.S. 269 (1942).
On the other hand, juror nullification also played a role in ugly events. During the Jim Crow and Civil Rights eras, nullification was “used to sanction murder and lynching.” Thomas, 116 F.3d at 616 (collecting sources). Moving from the ugly to merely the bad, we find cases in which nullification is used to protect “the defendant‘s shooting of his wife‘s paramour, or purchase during Prohibition of alcoholic beverages.” Dougherty, 473 F.2d at 1130.
The settled conclusion of American courts has been that nullification is a power but not a right. See, e.g., Thomas, 116 F.3d at 615. That is to say, jurors can vote to acquit in defiance of the law, and courts in general cannot stop them—but courts will not encourage this, provide jury instructions acknowledging it, or allow lawyers to argue overtly for it. See, e.g., United States v. Krzyske, 836 F.2d 1013, 1021 (6th Cir. 1988); Thomas, 116 F.3d at 617. But see Krzyske, 836 F.2d at 1021-22 (Merritt, J., dissenting); cf. United States v. Gabrion, 648 F.3d 307, 324 (6th Cir. 2011) (Merritt, J.), rehg en banc granted, opinion vacated (Nov. 17, 2011) on reh‘g en banc, 719 F.3d 511 (6th Cir. 2013). Therefore, when courts have crafted doctrines to protect against jurors’ being removed for their views on the case, they are not doing so to protect jury nullification.
We categorically reject the idea that, in a society committed to the rule of law, jury nullification is desirable or that courts may permit it to occur when it is within their authority to prevent. Accordingly, we conclude that a juror who intends to nullify the applicable law is no less subject to dismissal than is a juror who disregards the court‘s instructions due to an event or relationship that renders him biased or otherwise unable to render a fair and impartial verdict.
D. The Williams and Johnson Precedents
State-court decisions requiring an actual constitutional violation to overturn a conviction because of juror removal, federal decisions imposing a prophylactic rule, and AEDPA all came together in the case of Tara Williams, which eventually reached the Supreme Court in Johnson v. Williams, 568 U.S. 289 (2013). Because of the significant number of parallels between that case and ours, we discuss it in some detail. Williams was the getaway driver in a liquor-store robbery in which her coconspirators shot and killed a teller. She was tried in a California state court for murder. In response to two jury notes (delivered simultaneously) indicating that one juror “has expressed an intention to disregard the law and has expressed concern relative to the severity of the charge,” the judge interviewed each juror individually on the record. Williams v. Cavazos, 646 F.3d 626, 631–32 (9th Cir. 2011), rev‘d sub nom. Johnson v. Williams, 568 U.S. 289 (2013). In the face of more equivocal evidence than that which is before us today (including a split among the other jurors as to whether the juror in question was following the law and deliberating properly), the judge concluded that the juror should be dismissed as biased, citing evidence that suggested the juror disagreed with the law and had been dishonest. Id. at 634. An alternate was seated and Williams was convicted. Id. at 634-35.
Williams appealed on both state statutory and
Williams petitioned the federal courts for a writ of habeas corpus. The Ninth Circuit held that the California Court of Appeal had not “adjudicated Williams‘s
The Supreme Court did not agree. The Court began its analysis by noting its previous decision in Richter, in which it had held that when a federal claim has been presented to a state court and denied without explanation, a federal court should apply a (rebuttable) presumption “that the state court adjudicated the claim on the merits.” Johnson, 568 U.S. at 298. “We see no reason why the Richter presumption should not also apply when a state-court opinion addresses some but not all of a defendant‘s claims.” Ibid.30 This holding is, of course, important to our case.
Even more significant to us, however, is the Court‘s extended discussion of the way in which the Ninth Circuit‘s decision had failed to appreciate the import of the California Court of Appeal‘s discussion of the California Supreme Court precedent, Cleveland. Cleveland, the Supreme Court noted, had discussed but “expressly declined to follow” Brown, Thomas, and Symington. 568 U.S. at 304. The discussion of these cases, the Supreme Court observed, showed something more:
Cleveland did not expressly purport to decide a federal constitutional question,
but its discussion of Symington, Thomas, and Brown shows that the California Supreme Court understood itself to be deciding a question with federal
constitutional dimensions. Indeed, it is difficult to imagine the California Supreme Court announcing an interpretation of
. . . .
Regardless of whether a California court would consider Williams‘s § 1089 and
Sixth Amendment claims to be perfectly coextensive, the fact that these claims are so similar makes it unlikely that the California Court of Appeal decided one while overlooking the other. Indeed, it is difficult to imagine any panel of appellate judges reading Cleveland and passing on the propriety of dismissing a holdout juror under § 1089 without realizing that such situations also bear on the federal constitutional right to a fair trial. The California Court of Appeal‘s quotation of our definition of “impartiality” from [United States v. Wood, 299 U.S. 123, 145–146, (1936)], points to the same conclusion, confirming that the state court was well aware that the questioning and dismissal of [the juror] implicated both state and federal law.
568 U.S. at 305–06 (citations omitted). Wood was a
For these reasons, the Supreme Court held that that “the Ninth Circuit erred by finding that the California Court of Appeal overlooked Williams’
The Ninth Circuit‘s conclusion to the contrary rested on the fact that Cleveland refused to follow Symington, Brown, and Thomas. 646 F.3d at 640. But the views of the federal courts of appeals do not bind the California Supreme Court when it decides a federal constitutional question, and disagreeing with the lower federal courts is not the same as ignoring federal law. The Ninth Circuit‘s apparent assumption that the California Supreme Court could not refuse to follow federal court of appeals precedent without disregarding the Federal Constitution would undo
§ 2254(d) ‘s “contrary to” provision, which requires deference unless a state court fails to follow Supreme Court precedent.28 U.S.C. § 2254(d)(1) .
Id. at 305 (first emphasis added; second in original).
The parallels to our case seem painfully obvious. Though the district court and Wofford in our case have raised various grounds on which to differentiate the Williams cases from ours (see infra, pp. 37–41), the facts remain: (1) at trial, in both cases, a holdout juror was removed for cause that was related to, but distinct from, the juror‘s holdout status; (2) on direct appeal, the California Court of Appeal, like the MCOA, confronted a case in which defendant raised both state and federal claims on brief in a mixed fashion; (3) in each case the state appeals court‘s decision cited only state precedent; (4) on habeas, the Ninth Circuit, like our district court, held that a state appeals court had overlooked the
But an eight-Justice majority of the Supreme Court (with the ninth Justice writing an even more emphatic concurrence) reversed the Ninth Circuit. The Supreme Court held that it was error to presume that the California courts had ignored the constitutional question in Taylor and in Cleveland. Tracing references in these state-court decisions to federal cases on the
Given the significant similarities between Williams‘s case and Wofford‘s, we should do as the Supreme Court did and reverse—absent some other argument distinguishing the two. To such arguments we now turn.
E. Purported Distinguishing of Wofford from Williams and Johnson Fails
The district court and Wofford have both attempted to distinguish our case from Johnson on various grounds. Ultimately, none are convincing.
As we have seen, after Richter and Johnson, only one narrow path remained for those who would argue that a claim has been properly presented to the state court (so as to be preserved for habeas review and not forfeited under AEDPA), but not deemed to have been reviewed on the merits and decided by the state courts, even sub silentio: arguing that a claim had been presented, but fundamentally misunderstood. This does happen,31 and indeed the difference between the claim as presented and the claim as examined by the state courts can be fairly narrow and still overcome the post-Johnson AEDPA barriers to de novo review.32
The district court held that this is Wofford‘s situation. First, the district court held—correctly—that Wofford had raised a
In attempting to differentiate Johnson, the district court examined the MCOA‘s engagement with Tate, 624 N.W.2d 524. The district court emphasized certain language from Tate cited in the MCOA‘s Wofford opinion: “While a defendant has a fundamental interest in retaining the composition of the jury as originally chosen, he has an equally fundamental right to have a fair and impartial jury made up of persons able and willing to cooperate, a right that is protected by removing a juror unable or unwilling to cooperate.” Tate, 624 N.W.2d at 529 (emphasis added by district court). The district court observed that the emphasized language “stems from a line of federal cases holding that once a jury is empaneled, ... by declaring a mistrial too soon, a trial court may infringe the defendant‘s
Thus, this case is unlike Johnson v. Williams, 568 U.S. 289, 305 (2013), where the Supreme Court reversed the Ninth Circuit‘s de novo application of Symington. See 568 U.S. at 305 (finding
§ 2254(d) applied where state court cited Cleveland, and, in turn, Cleveland discussed Brown, Thomas, and Symington).
Thus, the district court‘s logic runs, because the Michigan courts cited a state-court line of reasoning on juror removal that derived from a
But this analysis is not convincing. First, the district court‘s analysis of the MCOA‘s silence as to Brown, Thomas, Symington, or the rule they embody is troubling. As we have just seen, the district court used this silence as a basis both for holding that the MCOA had overlooked Wofford‘s
The bigger problem is that the court‘s expressio unius analysis rests on a purported division between the
Just as the Supreme Court found it “difficult to imagine the California Supreme Court [in Cleveland] announc[ed] an interpretation of
Our conviction in reaching both of these conclusions comes not just from presumption; rather, it is reinforced significantly by Dry Land‘s discussion of
Incidentally, the history recited above also means that even on the district court‘s own logic, the attempted distinctions would fail. The connection that shows consideration of the
- CA Court of Appeal→Cleveland→Brown/Symington/Thomas→Sixth Amendment
- MI Court of Appeals→Tate→Dry Land→Phillips→Sixth Amendment
Whatever else the case may turn on, the constitutional analysis cannot turn on that extra level of citation. Attempts to differentiate Johnson v. Williams, in other words, have only strengthened the conviction that it should apply.33
The courts in the Brown-Thomas-Symington cases, just as every other court, agree that it would be constitutional to remove a juror who is misbehaving, ill, or otherwise unable to deliberate. See Brown, 823 F.2d at 597; Thomas, 116 F.3d at 613–14; Symington, 195 F.3d at 1085. What sets the Brown line apart from the Phillips-Dry Land line or Cleveland is that Brown is prophylactic. Reasoning that it will not always be easy to tell when a dismissal is justified versus when it stems from a juror‘s opinion on the case—especially without penetrating the secrecy of the jury room—each version of the Brown test requires that the dismissal be denied (or, in the appellate posture after conviction, that a new trial be ordered) in any case where there is a reasonable possibility that the juror‘s removal was related to the merits. See Brown, 823 F.2d at 596.
Cleveland is not prophylactic. It requires an actual constitutional violation to order a new trial. 21 P.3d at 1236. Dry Land also rejects a prophylactic approach, preferring to require actual prejudice when examining juror removal. And Johnson stands for the idea that which approach to take is a choice for the state to make.34
The background of Michigan law indicates that the Michigan courts are aware of
Admittedly, the MCOA considered the problem in light of the rules as distilled into one case, not with the explication displayed here today. But we are not here to grade Michigan‘s homework, only to assess whether there was an egregious constitutional violation. Johnson and Richter are reminders that the federal courts cannot “impose mandatory opinion-writing standards on state courts.” Johnson, 568 U.S. at 300; see also Richter, 562 U.S. at 99.
The conclusion from the foregoing is clear: in light of Johnson, the district court should have afforded AEDPA deference to the Michigan Wofford ruling, if not under Johnson‘s presumption, then because of the manifest similarities between Wofford and Johnson. (Indeed, the district court‘s repeated citations to the overturned Williams v. Cavazos decision are troubling: saying “rev‘d on other grounds sub nom. Johnson v. Williams” does not make it so.) The district court should, moreover, have recognized that, as a close look at Tate shows, the Michigan courts were drawing on a line of state precedent that considered the
V. Legal Analysis Under AEDPA, Applied
Wofford needs a prophylactic rule, because without one he must lose: the Michigan courts found—and we have upheld the finding—that Juror M was removed from the jury for cause. See supra, pp. 18–20. The MCOA ruled that “flagrantly violat[ing] the court‘s instructions by discussing the deliberations with a non-juror” provided good cause for dismissal under Michigan law. Wofford, 2015 WL 1214463 at *2. Indeed, Wofford did not argue before the MCOA, and does not argue before us, that the trial court lacked a just cause for Juror M‘s dismissal. Ibid. And the MCOA held that her dismissal was because of this conduct, not because of her opinion on the case. Id. at *2–3. We have upheld that factual finding as well. See supra, pp. 18–20. There was, in other words, no actual constitutional violation here—only some factors that might make a prophylactically minded court concerned about one. Michigan chose not to go in that direction.
Since AEDPA applies, we cannot use Brown-Thomas-Symington as the benchmark to evaluate the decisions of the Michigan courts. Johnson, 568 U.S. at 305. The Williams cases are yet again instructive: on remand from the Supreme Court, reconsidering the case under AEDPA, the Ninth Circuit found that there was no Supreme Court case directly on point and thus denied the petition for a writ of habeas corpus. Williams v. Johnson, 840 F.3d 1006, 1009–11 (9th Cir. 2016); see also id. at 1009 (“Williams has not cited any Supreme Court case imposing (or even hinting at) the Symington rule. Nor are we aware of such a case.“). Similarly, Wofford has not cited a Supreme Court case here that would support the application of the Brown-Thomas-Symington rule or of any other analogous rule. Therefore, Wofford‘s arguments
VI. Conclusion
For the foregoing reasons, we uphold the factual findings of the district court, but REVERSE its legal conclusion, and REMAND the case for further proceedings consistent with this opinion.