People of Michigan v. Daniel Albert Loew ( 2024 )

Michigan Supreme Court
Lansing, Michigan
Syllabus
Chief Justice: Justices:
Elizabeth T. Clement Brian K. Zahra
David F. Viviano
Richard H. Bernstein
Megan K. Cavanagh
Elizabeth M. Welch
Kyra H. Bolden
This syllabus constitutes no part of the opinion of the Court but has been Reporter of Decisions:
prepared by the Reporter of Decisions for the convenience of the reader. Kathryn L. Loomis
PEOPLE v LOEW
Docket No. 164133. Argued October 5, 2023 (Calendar No. 4). Decided July 16, 2024.
Daniel A. Loew was convicted following a jury trial in the Allegan Circuit Court of two
counts of first-degree criminal sexual conduct, MCL 750.520b(1)(f); one count of second-degree
criminal sexual conduct, MCL 750.520c(1)(f); and two counts of third-degree criminal sexual
conduct, MCL 750.520d(1)(a) and (b). Following his convictions, defendant appealed in the Court
of Appeals. Before his appeal was considered, defendant learned that the judge who presided over
the trial, Judge Margaret Zuzich Bakker (the trial judge), had corresponded via e-mail with Allegan
County’s elected prosecutor, Myrene Koch, during the trial. In the e-mails, the trial judge asked
questions about the investigation that led to the charges. Defendant moved for a new trial, alleging
judicial misconduct, ineffective assistance of counsel, and prosecutorial misconduct. On
defendant’s motion, the case was reassigned to a different trial court judge, Judge William A.
Baillargeon (the trial court), who granted defendant’s motion for a new trial on the basis that the
trial judge’s e-mails created the appearance of impropriety. The prosecution filed a cross-appeal
in the Court of Appeals contesting the decision to grant defendant’s motion for a new trial. The
Court of Appeals, MURRAY, P.J., and MARKEY, J. (RIORDAN, J., dissenting), reversed the order
granting defendant a new trial, holding that the trial court had abused its discretion by granting the
motion. 340 Mich App 100 (2022). Defendant sought leave to appeal in the Supreme Court,
which granted his application. 510 Mich 952 (2022).
In a lead opinion by Chief Justice CLEMENT, joined by Justices ZAHRA and VIVIANO, an
opinion concurring in part, dissenting in part, and concurring in the judgment by Justice BOLDEN,
and an opinion concurring in part and dissenting in part by Justice WELCH, joined by Justice
CAVANAGH, the Supreme Court held:
The trial judge’s conduct in this case violated the Michigan Code of Judicial Conduct. The
Court of Appeals erred by holding that the trial judge’s ex parte communications with Prosecutor
Koch were for administrative purposes under Canon 3(A)(4)(a) and therefore permissible. The
phrase “communications . . . for administrative purposes” means those communications made for
the purpose of managing or executing a pending or impending proceeding. Because the trial
judge’s ex parte communications with Prosecutor Koch were not made for the purpose of
managing or executing a pending or impending proceeding, they violated Canon 3(A)(4)(a). The
trial judge should have known that grounds for her disqualification might have existed under
MCR 2.003(C)(1)(b)(ii), given that her ex parte communications with Prosecutor Koch were
related to defendant’s trial and might have led one to reasonably question whether the trial judge
was interested in seeing the prosecution succeed or seeing defendant convicted. While these
communications did not show that the trial judge was actually biased or that there was an
unconstitutionally high probability she was actually biased, an ordinary person might still
reasonably question her impartiality. Therefore, under Canon 3(C), she should have raised the
issue of her disqualification sua sponte, and she should have recused herself. However, the ex
parte communications did not violate defendant’s constitutional right to trial by an impartial
decision-maker, right to be present, or right to counsel.
Affirmed.
In the lead opinion, Chief Justice CLEMENT, joined by Justices ZAHRA and VIVIANO, stated:
1. Even if due process does not require a judge to recuse themselves, MCR
2.003(C)(1)(b)(ii) may still so require if the judge, based on objective and reasonable perceptions,
has failed to adhere to the appearance of impropriety standard set forth in Canon 2 of the Michigan
Code of Judicial Conduct. Canon 2(A) states that a judge must avoid all impropriety and
appearance of impropriety. To decide whether a judge has failed to avoid the appearance of
impropriety requires a consideration of whether an ordinary person might reasonably question the
judge’s integrity, impartiality, or competence on the basis of the judge’s observable conduct.
Canon 3(A)(4) prohibits a judge from communicating with a party to a legal proceeding outside
the presence of opposing counsel in most instances. In particular, a judge may not initiate, permit,
or consider ex parte communications, but a judge may allow ex parte communications for
administrative purposes, as long as the judge reasonably believes that no party or counsel for a
party will gain a procedural or tactical advantage and the judge promptly discloses the
communication. In this case, the trial judge commented about the police investigation and
questioned whether the Michigan State Police had detectives and why the victim was not referred
for a medical examination. The Court of Appeals erred by concluding that these ex parte
communications with Prosecutor Koch were for administrative purposes. Communications for
administrative purposes are those communications made for the purpose of managing or executing
a pending or impending proceeding. Because the trial judge’s ex parte communications with
Prosecutor Koch were not made for this purpose, they violated Canon 3(A)(4)(a).
2. While these communications did not show actual bias or an unconstitutionally high
probability of bias, they might still have caused an ordinary person to reasonably question the
judge’s impartiality. No matter the content of the ex parte communications, it is a gross breach of
the appearance of justice when a party’s principal adversary is given private access to the ear of
the court. While a brief ex parte exchange concerning a matter unrelated to the defendant or the
proceeding might not create in reasonable minds a perception that the judge is biased, the trial
judge’s ex parte exchange with Prosecutor Koch—in which the trial judge expressed concern about
law enforcement’s missteps in its investigation of defendant’s case specifically and asking why
these missteps occurred—might lead one to reasonably question whether the trial judge was
interested in seeing the prosecution succeed or seeing defendant convicted. Not only did the trial
judge give Prosecutor Koch private access to her ear, considering the contents of her
communications, one might reasonably question whether she was interested in seeing the
prosecution succeed or seeing defendant convicted. For that reason, she should have known that
grounds for her disqualification might have existed under MCR 2.003(C)(1)(b)(ii). Under Canon
3(C), she should have raised the issue of her disqualification sua sponte, and she should have
recused herself.
3. Ex parte communications between a judge and the prosecution are not per se
unconstitutional. However, depending on the circumstances, ex parte communications might
deprive a defendant of the constitutional right to be present, the right to counsel, or the due-process
right to a fair trial more generally. Ex parte communications might also be evidence that a
defendant has been deprived of the right to a trial before an impartial judge, depending on the
nature, extent, and content of the communications. A defendant is deprived of the right to a trial
before an impartial judge only when the judge is actually biased or, if there is no evidence of actual
bias, when the probability of actual bias on the part of the judge or decision-maker is too high to
be constitutionally tolerable. Situations identified as presenting that risk include when the judge
or decision-maker (1) has a pecuniary interest in the outcome, (2) has been the target of personal
abuse or criticism from the party before the judge, (3) is enmeshed in other matters involving the
petitioner, or (4) might have prejudged the case because of prior participation as an accuser,
investigator, fact-finder, or initial decision-maker. There is a strong presumption of judicial
impartiality, and a party arguing otherwise bears a heavy burden to rebut this presumption. In this
case, the fact that the trial judge’s ex parte communications with Prosecutor Koch might have
caused one to reasonably question the trial judge’s impartiality did not inevitably show that she
was actually biased or that the appearance of bias was too high to be constitutionally tolerable.
While a judge’s having engaged in ex parte communications with the prosecution could show that
the judge is actually biased or that the appearance of bias was too high to be constitutionally
tolerable, this would depend on the character, substance, or extent of the ex parte communications
and must be examined on a case-by-case basis. In contrast to the factual situation in United States
v Barnwell, 477 F3d 844, 852 (CA 6, 2007), in this case, the trial judge and the prosecutor were
not extensively involved with each other or engaging in multiple lengthy meetings, the trial judge
did not cooperate with the prosecution on some matter in an effort to further the prosecution’s
interest to defendant’s detriment, and defendant identified nothing else to suggest that the trial
judge was actually biased or that she had a substantial incentive to be biased. The trial judge had
no pecuniary or personal interest in the outcome of this case, she had not been the target of personal
abuse or criticism from defendant, she had not been enmeshed in other matters involving
defendant, and she had not previously participated as an accuser, investigator, fact-finder, or initial
decision-maker in defendant’s case. Accordingly, defendant did not establish that he was deprived
of his due-process right to a trial presided over by an impartial judge and jury on the basis of the
ex parte communications at issue.
4. The trial judge’s ex parte communications did not deprive defendant of his right to be
present or his right to counsel. The Due Process Clauses and the Confrontation Clauses of the
Michigan Constitution and the United States Constitution impliedly guarantee defendants the right
to be present at any stage of a criminal proceeding in which their substantial rights might be
adversely affected. In addition, the Sixth Amendment and Const 1963, art 1, § 20, guarantee to
the accused the right to counsel at all critical stages of a criminal proceeding, which has likewise
been interpreted to mean that a defendant has the right to counsel at any stage at which a
defendant’s substantial rights may be affected. In this case, the interaction between the trial judge
and Prosecutor Koch was not one in which defendant’s substantial rights could have been affected.
The trial judge was not considering whether to take some action that would affect defendant, and
Prosecutor Koch was not suggesting or implying that the trial judge should take some action that
would do so. Instead, the trial judge and Prosecutor Koch had a brief, albeit inappropriate, side
conversation about law enforcement’s negligence, a topic of which defendant and defense counsel
were fully aware. Further, there was no indication that the trial judge’s ex parte communications
deprived defendant of due process because, in reaction to the trial judge’s comments to Prosecutor
Koch, the trial prosecutor altered her strategy. There was no evidence that Prosecutor Koch told
the trial prosecutor about the trial judge’s comments, and even if Prosecutor Koch’s knowledge
were imputed to the trial prosecutor, there was nothing to suggest that the trial prosecutor tried a
different tack in response to the trial judge’s comments. The trial prosecutor disclosed the
trooper’s investigative missteps in her opening statement, solicited testimony from the trooper
about his missteps, and solicited testimony about law enforcement’s failure to have the victim
undergo a medical examination, all of which occurred before the trial prosecutor could have known
about the trial judge’s ex parte communications. While defendant suggested that the trial
prosecutor spent more time in her closing statement emphasizing the trooper’s shortcomings than
she did in her opening statement, and that she mentioned the trooper’s investigative shortcomings
in her opening statement only to explain why the jury would not see any DNA evidence, that was
to be expected, given that opening statements are an attorney’s opportunity to give the jury an
overview of the evidence, not to make arguments about the defendant’s guilt or innocence. In
sum, there was nothing to suggest that the trial prosecutor altered her strategy in response to the
trial judge’s ex parte communications with Prosecutor Koch.
5. Although the trial judge should have recused herself under MCR 2.003(C)(1)(b)(ii), her
failure to do so did not result in a miscarriage of justice and did not give the trial court a legally
recognized basis for granting defendant a new trial under MCR 6.431(B). The failure to recuse
was a preserved, nonconstitutional error that required consideration of whether defendant was
entitled to relief under MCL 769.26, which restricts courts from granting a new trial in any criminal
case for error as to any matter of procedure only if the error resulted in a miscarriage of justice. A
nonconstitutional error results in a miscarriage of justice only if it had an effect on the finder of
fact to the defendant’s detriment. Typically, this requires the defendant to show that, but for the
error, it is more likely than not that the fact-finder would have reached a different conclusion. In
this case, the trial judge’s failure to recuse herself in violation of MCR 2.003(C)(1)(b) did not
result in a miscarriage of justice under MCL 769.26 because it had no effect on the finder of fact.
This was not a bench trial, and there was no dispute that the jury was unaware of the trial judge’s
ex parte communications. Therefore, though the trial judge should have recused herself under
MCR 2.003(C)(1)(b)(ii), that fact did not give the trial court a legally recognized basis on which
to grant defendant a new trial.
6. The standard articulated in Liljeberg v Health Servs Acquisition Corp, 486 US 847(1988), for determining whether a new trial is warranted was not adopted. Under Liljeberg, courts
consider the risk of injustice to the parties in the particular case, the risk that the denial of relief
will produce injustice in other cases, and the risk of undermining the public’s confidence in the
judicial process. The Liljeberg standard was not compatible with MCL 769.26 or Michigan
precedent, was vague to the point of futility, and would offer little guidance to courts in the exercise
of their discretion to grant or deny a new trial for a judge’s failure to recuse.
Justice BOLDEN, concurring in part, dissenting in part, and concurring in the judgment,
agreed with the lead opinion that the Court of Appeals judgment should be affirmed and agreed
that defendant failed to establish a miscarriage of justice based on the trial judge’s failure to recuse
on the basis of her ex parte communications. She also agreed with the majority’s decision not to
adopt the Liljeberg test given that it was not binding, was distinguishable because it involved a
statutory violation, and inadequately balanced the rights and interests at stake when applied in the
criminal context, particularly in light of its focus on efficiency. However, she would not have
concluded through these proceedings that the trial judge violated the Michigan Code of Judicial
Conduct because the focus of the Court’s analysis should have been limited to the crux of the issue,
which was whether defendant’s trial was fair.
Justice WELCH, joined by Justice CAVANAGH, concurring in part and dissenting in part, did
not join Part II(A)(3) of the lead opinion or the decision to affirm because she believed that the
Liljeberg framework could be applied consistently with harmless-error review under MCL 769.26
where the alleged error is inappropriate ex parte communications by a trial judge that create an
appearance of impropriety but do not rise to the level of a constitutional violation. Although
Justice WELCH did not believe that the ex parte communications at issue amounted to an error of
pleading or procedure for purposes of MCL 769.26, she accepted that the Court has long held that
MCL 769.26 controls the review of all preserved nonconstitutional errors and thus precedent
required application of the harmless-error standard provided by that statute. However, she opined
that the Liljeberg framework could be applied as mechanism for determining whether ex parte
communications resulted in a “miscarriage of justice” for purposes of MCL 769.26. Justice
WELCH also found instructive several decisions issued by the United States Court of Appeals for
the Seventh Circuit that had applied the Liljeberg framework in criminal cases involving similar
inappropriate ex parte communications that had been initiated by a presiding judge. She would
have remanded this case to a new trial court judge to examine the harmless-error standard through
the Liljeberg framework.
Justice BERNSTEIN, dissenting, agreed with the Court’s decision not to adopt the Liljeberg
test, and he joined the part of Justice BOLDEN’s opinion that explained why Liljeberg was an
unsuitable test for the Court to adopt. He also agreed with Justice BOLDEN that it was not necessary
to rely on the Michigan Code of Judicial Conduct to resolve the case. However, he would have
resolved the case on the constitutional question alone, addressing defendant’s argument solely in
due-process terms. He noted that the lead opinion’s approach, which puts the burden on defendants
to demonstrate a miscarriage of justice, was both unfair to defendants and a substantively futile
exercise in this context, given that ex parte communications and their effect on the prosecution are
all outside the trial court record, thus making a miscarriage of justice nearly impossible to prove.
Instead, Justice BERNSTEIN would have adopted the test that the United States Court of Appeals
for the Sixth Circuit set forth in Barnwell, which would have presumed that the ex parte
communications prejudiced defendant and put the burden on the prosecution to demonstrate that
the ex parte communications were justified by a compelling state interest and narrowly tailored
when possible. Because the parties did not have an opportunity to brief arguments under Barnwell,
he would have remanded this case to the Court of Appeals with instructions to apply the test and
determine whether defendant was entitled to a new trial.
Michigan Supreme Court
Lansing, Michigan
OPINION
Chief Justice: Justices:
Elizabeth T. Clement Brian K. Zahra
David F. Viviano
Richard H. Bernstein
Megan K. Cavanagh
Elizabeth M. Welch
Kyra H. Bolden
FILED July 16, 2024
STATE OF MICHIGAN
SUPREME COURT
PEOPLE OF THE STATE OF MICHIGAN,
Plaintiff-Appellee,
v No. 164133
DANIEL ALBERT LOEW,
Defendant-Appellant.
BEFORE THE ENTIRE BENCH
CLEMENT, C.J.
During defendant’s jury trial, the trial judge exchanged several e-mails with the
Allegan County Prosecutor regarding testimony of a Michigan State Police trooper and a
Michigan State Police detective. In her e-mails, the trial judge expressed concern about
mistakes law enforcement had made in its investigation and asked questions related to why
those mistakes had occurred. The trial judge never notified defendant or defense counsel
of these e-mails or their contents. It is clear that the trial judge’s private exchange with the
elected prosecutor violated the Michigan Code of Judicial Conduct. The issue here is
whether these ex parte communications warrant granting defendant a new trial under MCR
6.431(B). This issue has two components. The first component is whether the trial judge
should have recused herself under MCR 2.003(C)(1)(b)(ii), and if so, whether her failure
to do so resulted in a miscarriage of justice. The second component is whether defendant
was deprived of any constitutional rights because of the ex parte communications. For the
first component, we conclude that the trial judge should have recused herself under MCR
2.003(C)(1)(b)(ii) but that this did not result in a miscarriage of justice. For the second
component, we conclude that defendant was not deprived of any constitutional rights. We
therefore affirm the judgment of the Court of Appeals.
I. FACTS
Defendant was tried and convicted for repeatedly sexually assaulting the cousin of
his then fiancée, starting when the victim was 13 years old and he was 21 years old. The
jury convicted defendant of two counts of first-degree criminal sexual conduct (CSC-I), 1
one count of second-degree criminal sexual conduct (CSC-II), 2 and two counts of third-
degree criminal sexual conduct (CSC-III). 3 After defendant’s trial, he was sentenced as a
1
MCL 750.520b(1)(f) (defendant causes personal injury to the victim and uses force or
coercion).
2
MCL 750.520c(1)(f) (personal injury to victim and force or coercion).
3
One count of CSC-III was under MCL 750.520d(1)(a) (sexual penetration involving
victim at least 13 years of age and under 16 years of age), and the other was under MCL
750.520d(1)(b) (penetration by force or coercion).
2
third-offense habitual offender to concurrent prison terms of 20 to 40 years for the CSC-I
counts and 20 to 30 years for the CSC-II and CSC-III counts.
According to the victim, defendant sexually assaulted her multiple times in a
bathroom at her aunt’s house, and defendant would sometimes ejaculate on the bathmats
in the bathroom. After she reported the assaults to law enforcement in January 2018, she
was not referred for a medical examination, though this is usually done as a matter of
protocol. In addition, the Michigan State Police trooper tasked with responding to the
victim’s allegations did not collect the bathmats from the crime scene but rather from the
victim’s aunt, who delivered the bathmats to the trooper in a bag. The trooper never
showed the bathmats to the victim to confirm they were the correct bathmats, and he failed
to document the bathmats in the other bathrooms in the aunt’s house.
During her opening statement, the trial prosecutor told the jury about the trooper’s
deficient handling of the bathmats and that the bathmats did not have any DNA on them:
And we will hear, unfortunately, that there is no D.N.A. evidence.
[The victim] will testify that . . . she made her aunt aware, she made law
enforcement aware of blue bath mats that she last remembered the Defendant
ejaculating on. And you will hear from [the trooper] that [the victim’s] aunt
met him in the middle of the night at a gas station with a garbage bag full of
bath mats that were green, white, and blue.
Those bath mats were never taken and shown to the victim. Those
bath mats were not seized personally by law enforcement. But [the victim’s
aunt] turned those over and those obviously didn’t have any D.N.A. on them.
Later, the trial prosecutor called the trooper to testify, and the trooper recounted his
investigative mistakes: he admitted that he failed to properly collect the bathmats, that he
failed to show them to the victim, that he failed to properly photograph the crime scene,
3
that he failed to document the bathmats in the other bathrooms in the aunt’s house, and that
he failed to conduct timely and thorough witness interviews.
Just before the trial prosecutor finished questioning the trooper, the trial judge,
Judge Margaret Bakker, e-mailed the elected prosecutor for Allegan County, Myrene
Koch, stating: “This trooper didn’t do a very good investigation. Don’t they have
detectives with MSP anymore?” 4 Meanwhile, defendant’s jury trial continued. After
dismissing the trooper, the trial prosecutor called Michigan State Police Detective Sergeant
Todd Workman to the stand, who acknowledged that the victim had not been referred for
a medical examination, which contravened investigative protocol.
The next morning, just before the second day of trial began, Prosecutor Koch
responded to Judge Bakker’s e-mail asking about whether the Michigan State Police had
detectives. Prosecutor Koch replied: “They do but not typically for CSC’s. This trooper
has been given additional personal training since this investigation.” A few minutes later,
Judge Bakker responded, “One more question . . . this victim was not referred for a
medical, do you know why?” Judge Bakker then took the bench, and the second day of
trial began. Ten minutes into trial, Prosecutor Koch responded, “Yes, because the prior
APA assigned to the case did not catch that it was missed nor did anyone else who touched
the file. As a result, there will now be a checklist for CSC’s in files.” While the victim’s
older sister testified, Judge Bakker replied, “I thought Safe Harbor would catch it.” After
trial adjourned for the day, Prosecutor Koch responded, “Unfortunately, no. The forensic
interviewer is supposed to check that before case review but the list often is given to interns.
4
Throughout this opinion, Myrene Koch is referred to as “Prosecutor Koch” whereas “the
trial prosecutor” refers to the lead attorney who prosecuted defendant at trial.
4
I noticed it after the fact at case review but by then not clear on if the victim had much
support.”
During the trial prosecutor’s closing argument, the trial prosecutor addressed the
trooper’s admission that he had done a poor job investigating the alleged sexual assaults,
arguing as follows:
Do we know where these bath mats came from? Probably a bathroom,
I mean they are bath mats. Do we know, did they come from [defendant’s
fiancée’s] bathroom? Are they the light blue bath mats described to [the
trooper] the night that he first went to their home? No. Should [the trooper]
have opened this bag and looked at these bath mats and investigated it? Yes.
Should [the trooper] have . . . taken these bath mats, these blue, white, and
green bath mats to [the victim] and said, “Are these the bath mats you
remember being sexually assaulted on?” Yes.
If he didn’t do that, should he have taken a photograph of these bath
mats and sent that to [the victim] and said, “Are these green, white, light blue,
dark blue bath mats the blue bath mats you described?” Yes. But we didn’t
do that. Did he go that night to [the aunt’s] home where [defendant] resides
and take photographs of the inside of the home? You know, at this point,
[defendant] knows. So let’s just go ahead and deal with it. Should he have
gone and done that? Yes. He could have gone in and he could have seen the
bathroom that these . . . mats had been taken from. He could have seen a box
in storage that these . . . old mats had come from. But it didn’t happen. He
trusted the fact that this aunt was on [the victim’s] side. This aunt who has a
daughter who is now married to the Defendant. This aunt who has a
grandchild, the father of which is the Defendant. This aunt that has another
grandchild on the way, the father of which is the Defendant. Stacked the
cards in [defendant’s] favor. And we don’t have any way to verify where
these rugs come from.
All we’ve got is some lab reports that say that there are some blood
stains and one of those blood stains [defendant] is a major contributor of. I
am not saying they weren’t in the house. I am just saying these aren’t the
rugs that the victim described.
In the end, the jury convicted defendant, and defendant appealed by right.
Defendant was not aware of the trial judge’s ex parte communications at any point during
5
his trial or while he filed his direct appeal. But while defendant’s direct appeal was
pending, defendant learned of Judge Bakker’s ex parte communications with Prosecutor
Koch. In part on the basis of these ex parte communications, defendant moved for a new
trial, alleging that Judge Bakker violated Canon 3(A)(4) and Canon 2(A) of the Michigan
Code of Judicial Conduct and deprived him of due process of law. 5 After initially denying
defendant’s motion, Judge Bakker reassigned the case to a different judge. Judge William
A. Baillargeon granted defendant a new trial on the basis that Judge Bakker’s e-mails to
Prosecutor Koch created the appearance of impropriety in violation of Canon 2 and
deprived defendant of due process of law. 6
The prosecution cross-appealed, and the Court of Appeals reversed in a split,
published decision, holding that the trial court abused its discretion by granting defendant’s
motion for new trial. People v Loew, 340 Mich App 100, 104, 107-119; 985 NW2d 255(2022). The Court of Appeals first considered whether the trial judge had violated the
Michigan Code of Judicial Conduct. It held that the trial judge’s exchange with Prosecutor
Koch was allowed under Canon 3(A)(4) because the trial judge’s communications were for
administrative purposes, as the trial judge’s communications were about law enforcement’s
investigative process, not any matter relating to the merits. Still, the Court of Appeals
5
Defendant also argued that he was entitled to a new trial because defense counsel rendered
ineffective assistance of counsel and because the prosecutor committed misconduct by
eliciting perjured testimony. The trial court denied relief on those claims, and the Court of
Appeals similarly rejected them when defendant later raised them on appeal. Defendant
raised these issues again in his application for leave to appeal in this Court. We deny leave
to appeal as to those issues.
6
For the rest of this opinion, we will use “trial court” when referring to Judge Baillargeon
and “trial judge” when referring to Judge Bakker.
6
assumed for the sake of argument that the trial judge’s conduct gave the appearance of
impropriety in violation of Canon 2(A). Having found that the trial judge violated Canon
2(A), the Court of Appeals considered whether these violations deprived defendant of due
process of law. It concluded they did not because the trial judge’s ex parte communications
did not influence the jury in any way.
The dissenting Court of Appeals judge agreed that the trial judge’s ex parte
communications did not deprive defendant of due process of law but found that, because
the communications gave the appearance of impropriety in violation of Canon 2(A), the
trial judge should have recused herself under MCR 2.003(C)(1)(b)(ii). To decide whether
her failure to do so warranted granting a new trial, the dissent would have applied a
standard articulated by the United States Supreme Court in Liljeberg v Health Servs
Acquisition Corp, 486 US 847, 864; 108 S Ct 2194; 100 L Ed 2d 855 (1988). Under this
standard, the dissent concluded that the trial court did not abuse its discretion by granting
defendant a new trial.
Defendant sought leave to appeal here, arguing that the Court of Appeals erred by
concluding that the trial judge’s ex parte communications with Prosecutor Koch were
allowed under Canon 3(A)(4) and that her communications did not violate his
constitutional rights. In agreement with the Court of Appeals dissent, he also argues that
the trial judge’s communications gave the appearance of impropriety in violation of Canon
2, that the trial judge should have recused herself under MCR 2.003(C)(1)(b)(ii), and that
her failure to do so requires reversal under the Liljeberg standard. We granted leave to
appeal, limited to the issues of judicial misconduct alleged in Issue II of defendant’s
application. People v Loew, 510 Mich 952 (2022).
7
II. ANALYSIS
Under MCR 6.431(B), a trial court “may order a new trial on any ground that would
support appellate reversal of the conviction or because it believes that the verdict has
resulted in a miscarriage of justice.” 7 We review a trial court’s decision on a motion for
new trial for an abuse of discretion. Kailimai v Firestone Tire & Rubber Co, 398 Mich
230, 232; 247 NW2d 295 (1976). A trial court abuses its discretion if it grants a new trial
without providing a legally recognized basis for relief or if its basis for relief rests on an
unreasonable interpretation of the record. See id. at 233.
Initially, we note that a judge’s violation of the Michigan Code of Judicial Conduct
is not a legally recognized basis for relief. We implemented the Michigan Code of Judicial
Conduct in 1974 to provide ethical guidance to jurists in the discharge of their duties. See
Attorney General v Pub Serv Comm, 243 Mich App 487, 492; 625 NW2d 16 (2000). A
judge’s violation of a canon may be grounds for us to exercise our power to discipline that
judge, see MCR 9.202(B)(2); Const 1963, art 6, § 30, but the canons do not grant litigants
any substantive or procedural rights. Accord Goodheart v Casey, 523 Pa 188, 198; 565
A2d 757 (1989) (holding that a state code of judicial conduct provides standards for judges
to assess their own conduct and does not confer substantive rights on the parties to the
7
In the same vein, MCL 770.1 states that a court “may grant a new trial to the defendant,
for any cause for which by law a new trial may be granted, or when it appears to the court
that justice has not been done, and on the terms or conditions as the court directs.”
Defendant’s motion for a new trial relied on MCR 6.431(B) and not MCL 770.1. We leave
for another day the question of whether, and to what extent, any substantive difference
exists between MCR 6.431(B) and MCL 770.1. See People v Herbert, 444 Mich 466, 475
n 12; 511 NW2d 654 (1993) (noting the potential conflict between MCR 6.431(B) and
MCL 770.1 but declining to address whether there is a “substantive difference” between
the two), overruled in part on other grounds by People v Lemmon, 456 Mich 625 (1998).
8
litigation in question). Therefore, to be entitled to a new trial under MCR 6.431(B), a
defendant must do more than show that a judge violated the Michigan Code of Judicial
Conduct.
Aside from the trial judge’s violations of the Michigan Code of Judicial Conduct,
defendant has named two legal bases for relief in this case. We address each of these
arguments in turn. First, defendant argues that the trial judge violated MCR
2.003(C)(1)(b)(ii) by not recusing herself for failing to adhere to the appearance-of-
impropriety standard in Canon 2(A). This issue does not implicate defendant’s
constitutional rights and is therefore analyzed as a nonconstitutional error. Second,
defendant argues that the trial judge’s ex parte communications with Prosecutor Koch
violated his right to due process of law, his right to be present, and his right to counsel.
These questions require us to consider whether defendant was deprived of his right to a fair
trial, a constitutional question of due process. In re Murchison, 349 US 133, 136; 75 S Ct
623; 99 L Ed 942 (1955) (“A fair trial in a fair tribunal is a basic requirement of due process.
Fairness of course requires an absence of actual bias in the trial of cases.”). Due process
does not require a judge to recuse herself unless a judge is actually biased, Bracy v
Gramley, 520 US 899, 904-905; 117 S Ct 1793; 138 L Ed 2d 97 (1997), or, if there is no
evidence that the judge is actually biased, unless the situation is one in which “ ‘experience
teaches that the probability of actual bias on the part of the judge or decisionmaker is too
high to be constitutionally tolerable,’ ” Crampton v Dep’t of State, 395 Mich 347, 351; 235
NW2d 352 (1975), quoting Withrow v Larkin, 421 US 35, 47; 95 S Ct 1456; 43 L Ed 2d
712 (1975).
9
A. MCR 2.003(C)(1)
Defendant’s constitutional concerns are addressed later in this opinion. But to
begin, even if due process does not require a judge to recuse herself, MCR
2.003(C)(1)(b)(ii) may still require a judge to disqualify herself if “[t]he judge, based on
objective and reasonable perceptions, . . . has failed to adhere to the appearance of
impropriety standard set forth in Canon 2 of the Michigan Code of Judicial Conduct.”
Canon 2(A) states that “[a] judge must avoid all impropriety and appearance of
impropriety.” To decide whether a judge has failed to avoid the appearance of impropriety,
we consider whether the judge’s conduct “ ‘would create in reasonable minds a perception
that the judge’s ability to carry out judicial responsibilities with integrity, impartiality and
competence is impaired.’ ” People v Aceval, 486 Mich 887, 889 (2010) (statement of
HATHAWAY, J.), quoting Caperton v AT Massey Coal Co, Inc, 556 US 868, 888; 129 S Ct
2252; 173 L Ed 2d 1208 (2009) (additional quotation marks and citation omitted). Simply
put, we consider whether an ordinary person might reasonably question the judge’s
integrity, impartiality, or competence on the basis of the judge’s observable conduct.
1. CANON 3(A)(4)—EX PARTE COMMUNICATIONS
Before considering whether the trial judge failed to adhere to the appearance-of-
impropriety standard set forth in Canon 2(A), we first consider whether she violated Canon
3(A)(4), because defendant argues that a violation of Canon 3(A)(4) is the basis for a
finding of an appearance of impropriety. 8 Canon 3(A)(4) prohibits a judge from
8
We pause to note that in In re Haley, 476 Mich 180, 194-195; 720 NW2d 246 (2006), we
held that if a judge’s conduct is either expressly permitted or expressly forbidden by
another specific canon, then we will not independently examine whether the judge’s
conduct also gave rise to an appearance of impropriety under Canon 2(A). But we reached
that holding in the context of deciding whether we should sanction a judge for violating
10
communicating with a party to a legal proceeding outside the presence of opposing counsel
in most instances. Canon 3(A)(4) provides:
A judge shall not initiate, permit, or consider ex parte
communications, or consider other communications made to the judge
outside the presence of the parties concerning a pending or impending
proceeding, except as follows:
(a) A judge may allow ex parte communications for scheduling,
administrative purposes, or emergencies that do not deal with substantive
matters or issues on the merits, provided:
(i) the judge reasonably believes that no party or counsel for a party
will gain a procedural or tactical advantage as a result of the ex parte
communication, and
(ii) the judge makes provision promptly to notify all other parties and
counsel for parties of the substance of the ex parte communication and allows
an opportunity to respond. [Emphasis added.]
In a word, a judge may not initiate, permit, or consider ex parte communications, but a
judge “may allow” ex parte communications “for administrative purposes,” so long as the
Canon 2 when Canon 5(C) already expressly prohibited his conduct. See id. at 183, 189,
194-195. Put otherwise, if a judge’s conduct is either expressly permitted or expressly
forbidden by another specific canon, we will not sanction the judge for violating Canon
2(A). But this does not mean we will not consider whether the judge’s conduct required
recusal under MCR 2.003(C)(1)(b)(ii). Hence, even if Canon 3(A) expressly controls the
trial judge’s conduct, as the Court of Appeals concluded, we still review whether the trial
judge’s conduct gave rise to the appearance of impropriety such that she should have
recused herself.
That said, we acknowledge that we could decide whether the trial judge failed to
adhere to the appearance-of-impropriety standard without regard to whether Canon 3(A)(4)
forbade or allowed her ex parte communications. But because we believe the Court of
Appeals erred by holding that Canon 3(A)(4)(a) allowed the trial judge’s ex parte
communications, and because we believe the trial judge’s violation of this canon is relevant
to deciding whether she failed to adhere to the appearance-of-impropriety standard, we first
explain why the trial judge’s ex parte communications were prohibited under Canon 3(A).
11
judge reasonably believes that no party or counsel for a party will gain a procedural or
tactical advantage and the judge promptly discloses the communication.
The Court of Appeals held that the trial judge’s ex parte communications with
Prosecutor Koch were for administrative purposes because they concerned law
enforcement’s investigative process and were not related to the merits. We conclude that
the Court of Appeals erred on this point.
To begin, we note that Canon 3(A)(4)(a) provides that a judge “may allow” ex parte
communications for administrative purposes, and we are skeptical that this means a judge
may initiate ex parte communications for administrative purposes. But that aside, the trial
judge commenting about the trooper’s investigation, asking whether the Michigan State
Police has detectives, and asking why the victim was not referred for a medical examination
were not “communications . . . for administrative purposes,” at least not as that phrase is
used in Canon 3(A)(4)(a). Divorced from context perhaps, the phrase
“communications . . . for administrative purposes” could plausibly refer to any
communication made for the purpose of managing or supervising the process of something,
no matter what that something is. 9 But this phrase appears in the context of a judicial canon
regulating a judge’s conduct in the performance of her adjudicative responsibilities. See
Dep’t of Talent & Econ Dev/Unemployment Ins Agency v Great Oaks Country Club, Inc,
507 Mich 212, 227; 968 NW2d 336 (2021) (noting that plain and ordinary meaning must
account for the context in which the words are used). It is clear, then, that
9
“Administrative” means “of or relating to administration or an administration:
EXECUTIVE,” and “administration” is best defined here as the “performance of executive
duties: MANAGEMENT” or “the act or process of administering.” Merriam-Webster’s
Collegiate Dictionary (11th ed).
12
“communications . . . for administrative purposes” means those communications made for
the purpose of managing or executing a pending or impending proceeding. Because the
trial judge’s ex parte communications with Prosecutor Koch were not made for the purpose
of managing or executing a pending or impending proceeding, they violated Canon
3(A)(4)(a).
2. REASONABLE PERCEPTION OF BIAS
With that in mind, we now consider whether an ordinary person might reasonably
question the trial judge’s integrity, impartiality, or competence on the basis of her private
exchange with Prosecutor Koch. As discussed later in Part II(B) of this opinion, while the
trial judge’s communications do not show she was actually biased or that there was an
unconstitutionally high probability she was actually biased, we conclude that an ordinary
person might still reasonably question her impartiality. No matter the content of the ex
parte communications, it is “a gross breach of the appearance of justice when [a party’s]
principal adversary is given private access to the ear of the court . . . .” United States v
Minsky, 963 F2d 870, 874 (CA 6, 1992) (quotation marks and citation omitted).
This is not to suggest that one instance of ex parte communications always requires
a judge to disqualify herself. See State v Lotter, 255 Neb 456, 475; 586 NW2d 591 (1998)
(“[A] trial judge must recuse himself or herself only when the ex parte communication
poses a threat to the judge’s impartiality.”), mod on den of reh 255 Neb 889 (1999).
Depending on the circumstances, a brief ex parte exchange concerning a matter unrelated
to the defendant or the proceeding might not create in reasonable minds a perception that
the judge is biased. But the trial judge’s ex parte exchange with Prosecutor Koch was not
13
about some matter unrelated to defendant or his trial. In response to witness testimony,
while presiding over defendant’s trial, the trial judge privately e-mailed Prosecutor Koch
expressing concern about law enforcement’s missteps in its investigation of defendant’s
case specifically and asking why these missteps occurred. Not only did the trial judge give
Prosecutor Koch “private access to” her ear, considering the contents of her
communications, one might reasonably question whether the trial judge was interested in
seeing the prosecution succeed or seeing defendant convicted. See Minsky, 963 F2d at 874.
For that reason, the trial judge should have known that grounds for her disqualification
might have existed under MCR 2.003(C)(1)(b)(ii). Under Canon 3(C), she should have
raised the issue of her disqualification sua sponte, and she should have recused herself.
3. MISCARRIAGE OF JUSTICE
Nonetheless, this is not enough to conclude that the trial court had a legally
recognized basis for granting defendant a new trial under MCR 6.431(B). This preserved,
nonconstitutional error requires us to consider whether defendant was entitled to relief
under MCL 769.26. 10 See People v Lukity, 460 Mich 484, 495; 596 NW2d 607 (1999);
10
To preserve for appellate review the argument that a judge should have recused herself,
typically a party must timely move for the judge’s recusal in compliance with MCR
2.003(D). See People v Bettistea, 173 Mich App 106, 123; 434 NW2d 138 (1988); Law
Offices of Lawrence J Stockler, PC v Rose, 174 Mich App 14, 23; 436 NW2d 70 (1989).
Notably, MCR 2.003(D)(1)(a) requires a party to move for disqualification “within 14 days
of the discovery of the grounds for disqualification.” If party has failed to do so, the issue
is unpreserved, and appellate courts typically apply the plain-error standard of review. See
People v Jones, 468 Mich 345, 355; 662 NW2d 376 (2003). Defendant did not timely
move to disqualify the trial judge during the trial, but given that the trial judge did not
disclose the grounds for her disqualification to defendant, he could not have been expected
to do so. See Commonwealth v Malovich, 903 A2d 1247, 1252; 2006 PA Super 183 (2006)
(holding that a defendant had not waived claims on appeal when he was unaware that the
claims needed to be preserved by a motion for reconsideration because of the trial court’s
14
People v Mateo, 453 Mich 203, 210-211; 551 NW2d 891 (1996). MCL 769.26 restricts
courts from granting a new trial in any criminal case for error as to any matter of procedure
“unless in the opinion of the court, after an examination of the entire cause, it shall
affirmatively appear that the error complained of has resulted in a miscarriage of justice.”
Therefore, defendant was entitled to a new trial because of this error only if the error
resulted in a “miscarriage of justice.”
Our cases agree that a nonconstitutional error results in a miscarriage of justice only
if a nonconstitutional error had an effect on the finder of fact to the defendant’s detriment.
See Lukity, 460 Mich at 492; Mateo, 453 Mich at 221 (“The defendant’s right to a fair trial
by jury requires that preserved error be reviewed in terms of its effect on the factfinder.”).
Typically, this requires the defendant to show that, but for the error, it is more likely than
not that the jury would have reached a different conclusion, Lukity, 460 Mich at 495-496,
though we have recognized that there are some nonconstitutional errors for which we
presume the error affected the finder of fact to the defendant’s detriment, see People v
Yarbrough, 511 Mich 252, 267-272; 999 NW2d 372 (2023) (holding that a trial court’s
erroneous denial of a peremptory challenge under the court rules requires automatic
reversal). Either way, the defendant must at minimum show that there was an error that
failure to inform him of his right to file postsentencing motions). Presuming, without
deciding, that traditional preservation doctrines apply here, when defendant had no
knowledge of the ex parte communications or to object to them during trial, given the
exceptional circumstances of this case, we will treat this issue as preserved and will not
apply the plain-error standard. See Booth Newspapers, Inc v Univ of Mich Bd of Regents,
444 Mich 211, 234 n 23; 507 NW2d 422 (1993) (noting that this Court has overlooked
failure to preserve issues under “exceptional circumstances”).
15
affected the finder of fact, even if in some instances we will presume the effect was
prejudicial.
In this case, the trial judge’s failure to recuse herself in violation of MCR
2.003(C)(1)(b) did not result in a miscarriage of justice under MCL 769.26 because it had
no effect on the finder of fact. This was not a bench trial, and there is no dispute that the
jury was unaware of the trial judge’s ex parte communications. Cf. People v Stevens, 498
Mich 162, 171; 869 NW2d 233 (2015) (holding that a judge’s misconduct in front of the
jury deprives a defendant of the constitutional right to a fair trial “when, considering the
totality of the circumstances, it is reasonably likely that the judge’s conduct improperly
influenced the jury by creating the appearance of advocacy or partiality against a party”).
Therefore, though the trial judge should have recused under MCR 2.003(C)(1)(b)(ii), the
trial court did not have a legally recognized basis on which to grant defendant a new trial
because of it.
Defendant urges us to adopt the view of Court of Appeals dissent and to apply the
standard articulated in Liljeberg to determine whether the trial court had a legal basis to
grant defendant a new trial. See Liljeberg, 486 US at 864. In Liljeberg, the United States
Supreme Court considered under what circumstances a federal court could vacate a
judgment under FR Civ P 60(b) on the basis of a judge’s failure to recuse herself under 28
USC 455a. 11 See id. at 862-864. The Court observed that it had previously held that FR
Civ P 60(b) allowed a federal court to relieve a party from final judgment whenever
“appropriate to accomplish justice” but had also cautioned that relief should be granted
11
Similar to MCR 2.003(C)(1), 28 USC 455(a) requires a judge to “disqualify himself in
any proceeding in which his impartiality might reasonably be questioned.”
16
only in “extraordinary circumstances.” Id. at 864 (quotation marks and citation omitted).
In light of this, the Court held that, to determine whether to vacate a judgment under FR
Civ P 60(b) for a judge’s failure to recuse herself in violation of 28 USC 455(a), courts
may “consider the risk of injustice to the parties in the particular case, the risk that the
denial of relief will produce injustice in other cases, and the risk of undermining the
public’s confidence in the judicial process.” Id. Defendant would have us consider the
same to decide whether Judge Bakker’s failure to recuse herself warrants a new trial.
We decline the invitation to do so. As just discussed, under MCL 769.26 and our
precedent interpreting it, a miscarriage of justice occurs only when a nonconstitutional
error affected the finder of fact. If a nonconstitutional error had no effect on the finder of
fact, a court’s inquiry under MCL 769.26 is at its end. Courts may not go on to consider
the concerns addressed in Liljeberg, including whether allowing the verdict to stand would
otherwise risk injustice to the parties, risk injustice to parties in future cases, or risk
undermining the public’s confidence in the judicial process. Besides, even if the Liljeberg
standard were compatible with MCL 769.26 and our precedent, we would be reluctant to
adopt it. In our view, the Liljeberg standard is vague to the point of futility. The Liljeberg
standard emphasizes notions of fairness and equity that in practice are difficult, if not
impossible, to incorporate in a consistent, rule-based manner. The notion of “injustice” is
broad enough to encompass a limitless range of considerations, and so it can always be
said that failure to grant a new trial would be unjust to someone for some reason. Likewise,
there is always some risk of undermining public confidence in the judicial process if a new
trial is not granted when a judge for whom there was reason to question her partiality
presided over the first. In brief, we believe the Liljeberg standard would offer little
17
guidance to courts in the exercise of their discretion to grant or deny a new trial for a judge’s
failure to recuse herself.
Defendant’s first proffered legal basis for relief therefore fails. Although the trial
judge would have been required to recuse herself under MCR 2.003(C)(1)(b)(ii) on
defendant’s motion, there was no legal basis under MCL 769.26 for granting defendant a
new trial on this ground, and we decline to adopt the Liljeberg standard.
B. CONSTITUTIONAL RIGHTS
We turn next to defendant’s second proffered legal basis for relief—that the trial
judge’s ex parte communications with Prosecutor Koch deprived him of his right to due
process, his right to be present, and his right to counsel. To start, we note that the mere
occurrence of an ex parte conversation between a judge and the prosecution, alone, does
not automatically deprive a defendant of any constitutional right. Cf. United States v
Gagnon, 470 US 522, 526; 105 S Ct 1482; 84 L Ed 2d 486 (1985) (“The mere occurrence
of an ex parte conversation between a trial judge and a juror does not constitute a
deprivation of any constitutional right.”) (cleaned up). 12 In other words, ex parte
communications between a judge and the prosecution are not per se unconstitutional. But
depending on the circumstances, ex parte communications between a judge and the
prosecution might deprive a defendant of the constitutional right to be present, to effective
assistance of counsel, or the due-process right to a fair trial more generally. See United
12
We recognize that Gagnon concerned ex parte conversations between a judge and juror,
whereas the ex parte communications here were between the trial judge and the
prosecution. Given that no fact-finders were aware of these ex parte communications, we
see no reason to provide greater constitutional rights in this situation than would have been
recognized in Gagnon.
18
States v Barnwell, 477 F3d 844, 852 (CA 6, 2007); Jones v State, 668 P2d 1170, 1171-
1172; 1983 OK CR 127 (Okla Crim App, 1983); Lotter, 255 Neb at 480-482. Ex parte
communications might also show that a defendant has been deprived of the right to a trial
before an impartial judge, depending on the character, substance, and extent of the
communications. See, e.g., Barnwell, 477 F3d at 852.
1. RIGHT TO TRIAL BY IMPARTIAL DECISION-MAKER
Defendant first suggests that he was deprived of his due-process right to a trial
presided over by an impartial judge and jury. We disagree.
The federal Due Process Clause of the Fourteenth Amendment and the Due Process
Clause of the Michigan Constitution, Const 1963, art 1, § 17, guarantee a trial before an
impartial judge and jury. See Marshall v Jerrico, Inc, 446 US 238, 242; 100 S Ct 1610;
64 L Ed 2d 182 (1980). 13 As noted before, a defendant is deprived of his right to a trial
before an impartial judge only when the judge is actually biased, Bracy, 520 US at 904-
905, or, if there is no evidence that the judge is actually biased, when “ ‘experience teaches
that the probability of actual bias on the part of the judge or decisionmaker is too high to
be constitutionally tolerable,’ ” Crampton, 395 Mich at 351, quoting Withrow, 421 US at
47. Situations identified by this Court as presenting that risk include when the judge or
decision-maker:
13
Although this Court is not bound by federal precedent interpreting the Fourteenth
Amendment’s Due Process Clause when interpreting the Michigan Constitution’s Due
Process Clause, because the two clauses are textually similar, we may still find United
States Supreme Court precedent interpreting the federal Due Process Clause persuasive.
See Bauserman v Unemployment Ins Agency, 503 Mich 169, 185 n 12; 931 NW2d 539(2019).
19
(1) has a pecuniary interest in the outcome; (2) has been the target of personal
abuse or criticism from the party before him; (3) is enmeshed in other matters
involving petitioner; or (4) might have prejudged the case because of prior
participation as an accuser, investigator, fact finder or initial decisionmaker.
[Cain v Dep’t of Corrections, 451 Mich 470, 498; 548 NW2d 210 (1996)
(cleaned up).]
There is a strong presumption of judicial impartiality, and a party arguing otherwise bears
a heavy burden to rebut this presumption. Mitchell v Mitchell, 296 Mich App 513, 523;
823 NW2d 153 (2012); see also Cain, 451 Mich at 498 (noting that “disqualification for
bias or prejudice is only constitutionally required in the most extreme cases”).
A judge’s engaging in ex parte communications with the prosecution may no doubt
undermine confidence in the impartiality of the court. Grievance Administrator v Lopatin,
462 Mich 235, 262-263; 612 NW2d 120 (2000), citing Shaman, Lubet & Alfini, Judicial
Conduct and Ethics (3d ed), § 5.01, pp 159-160; United States v Earley, 746 F2d 412, 416
(CA 8, 1984) (explaining that, no matter the court’s motives in doing so, “the practice
should be discouraged since it undermines confidence in the impartiality of the court”)
(quotation marks and citation omitted). Indeed, we acknowledged that the trial judge’s ex
parte communications with Prosecutor Koch might have caused one to reasonably question
the trial judge’s impartiality. Even so, while a trial judge engaging in ex parte
communications with the prosecution may give the appearance of bias, it does not
inevitably show that the trial judge was actually biased or that the appearance of bias was
too high to be constitutionally tolerable. See United States v Williams, 949 F3d 1056, 1061
(CA 7, 2020) (“[B]ad appearances alone do not require disqualification.”) (quotation marks
and citation omitted). While a judge’s having engaged in ex parte communications with
the prosecution could show that the judge is actually biased or that the appearance of bias
20
is too high to be constitutionally tolerable, this would depend on the character, substance,
or extent of the ex parte communications and must be examined on a case-by-case basis.
Compare Williams, 949 F3d at 1061-1063, with Barnwell, 477 F3d at 850-854.
The United States Court of Appeals for the Sixth Circuit’s decision in Barnwell
illustrates when a judge’s ex parte communications with the prosecution are enough to
show actual bias or that the appearance of bias is too high to be constitutionally tolerable. 14
14
Citing Minsky, 963 F2d at 874, the Sixth Circuit in Barnwell said that “[a]n ex parte
communication between the prosecution and the trial judge can only be ‘justified and
allowed by compelling state interest.’ ” Barnwell, 477 F3d at 850, quoting Minsky, 963
F2d at 874 (additional quotation marks and citation omitted). Relying on this passage,
Justice BERNSTEIN says he would adopt this rule and hold that ex parte communications
between a judge and the prosecution about a substantive matter violate a defendant’s due-
process rights unless the prosecution can show that there was a compelling state interest
for the communication.
In our view, Justice BERNSTEIN misinterprets Barnwell and Minsky. In Minsky, the
trial court held an ex parte meeting with the prosecution to decide whether the prosecution
had to disclose to the defense a certain report containing potential impeachment evidence
at a stage where the defense was arguing that the report was “subject to disclosure under
the Jencks Act and Brady [v Maryland, 373 US 83; 83 S Ct 1194; 10 L Ed 2d 215 (1963),]”
and the release of the report would have allowed the defense to undermine the credibility
of a key government witness. Minsky, 963 F2d at 874. In Barnwell, the trial court held
multiple ex parte meetings with the prosecution during jury deliberations to discuss how to
handle a potentially biased juror who was brought to light through a preexisting
government wiretap related to other individuals who resided in the same county as the
defendant. See Barnwell, 477 F3d at 847-849. Hence, in both Barnwell and Minsky, the
ex parte interaction was one initiated by the prosecutor and was an interaction in which the
defendant’s substantive rights were at stake, and so the defendant’s exclusion from the
interaction violated his due-process right to be present at a critical stage of the proceeding.
This right is no doubt a fundamental one. See Washington v Glucksberg, 521 US 702, 720-
721; 117 S Ct 2258; 138 L Ed 2d 772 (1997) (“[T]he Due Process Clause specially protects
those fundamental rights and liberties which are, objectively, deeply rooted in this Nation’s
history and tradition . . . .”) (quotation marks and citation omitted). That said, the Sixth
Circuit in Barnwell and Minsky was simply applying strict scrutiny to decide whether the
government’s infringement on this fundamental right was justified. See Does v Munoz,
507 F3d 961, 964 (CA 6, 2007) (“Government actions that burden the exercise
of . . . fundamental rights or liberty interests are subject to strict scrutiny, and will be
21
In Barnwell, assistant United States attorneys and the judge had five ex parte meetings
during jury deliberations to discuss a juror who might have been partial and to discuss
potential jury tampering. See Barnwell, 477 F3d at 847-849. The assistant United States
attorneys and the judge did not include defense counsel in these conversations because the
United States Attorney’s Office learned that the juror might be partial from an FBI wiretap
in an unrelated investigation, and it was worried that disclosing this fact might compromise
the wiretap in the unrelated investigation. See id. at 851. The judge gave the assistant
United States attorneys an open invitation to contact and meet with him about the matter
as much as they deemed necessary. Id. At one of their meetings, the judge told the assistant
United States attorneys that they would “handle the matter ‘jointly,’ ” and in another
meeting he offered suggestions about what he could do to assist the government with its
investigation into the juror. See id. Further, the record showed that the judge took actions
during the proceedings to promote the prosecution’s interests to the defendant’s detriment:
(1) the judge sent the jury to lunch apparently to help the government catch the possibly
partial juror in the act of talking about deliberations, (2) the judge interviewed the jury
foreperson off the record over the defendant’s objection, and (3) after the jury hung, the
judge denied defense counsel’s request to poll the jury and handled other defense requests
upheld only when they are narrowly tailored to a compelling governmental interest.”)
(quotation marks and citation omitted). As we discuss later in this opinion, the interaction
between the trial judge and Prosecutor Koch was not one in which defendant’s substantive
rights were at stake, so his exclusion from this interaction did not violate his due-process
right to be present. For that reason, we need not decide whether the government had a
compelling interest for the communication. We do not read Barnwell or Minsky as holding
that any ex parte communication between the prosecution and a judge that is substantive
in nature violates a defendant’s right to due process unless the government has a
compelling interest for the communication.
22
that would have allowed the jury to make a decision “in short order.” Id. at 851-852. Under
these circumstances, the Sixth Circuit held that these ex parte communications deprived
the defendant of his right to a trial presided over by an impartial judge. See id. at 850. 15
The facts here are poles apart from Barnwell. To begin, unlike in Barnwell, where
members of the prosecution and the judge had multiple meetings, the trial judge sent only
three short e-mails to Prosecutor Koch asking questions about the investigation in
defendant’s case. This was not a case in which Judge Bakker and Prosecutor Koch were
extensively involved with each other or engaging in lengthy meetings. More importantly,
unlike in Barnwell, the trial judge did not cooperate with the prosecution on some matter
in an effort to further the prosecution’s interest to defendant’s detriment. The trial judge
sent a few brief e-mails expressing her concern about the sufficiency of law enforcement’s
investigation. Finally, the judge in Barnwell made multiple comments suggesting that he
viewed himself as being on the same team as the prosecution: he gave the prosecution an
open invitation to meet with him as much as necessary, he assured the prosecution that he
would “handle the matter jointly” with them, and he offered suggestions about what he
could do to help the prosecution with its investigation. Here, the trial judge made no such
comments. She did not say anything to suggest that she thought the possible deficiencies
in the trooper’s or Detective Workman’s investigative conduct posed a problem for the
prosecution’s case, and she and Prosecutor Koch did not discuss anything—positive or
negative—related to the trial prosecutor’s trial strategy.
15
The Sixth Circuit also held that the judge’s ex parte communications deprived defendant
of his right to be present and his right to effective assistance of counsel. See id.
23
Altogether, the trial judge’s ex parte communications here were not of such a
character, substance, or extent as to suggest that the trial judge was actually biased or that
the probability she was actually biased was too high to be constitutionally tolerable. And
apart from these ex parte communications, defendant has identified nothing else to suggest
that the trial judge was actually biased or that she had a substantial incentive to be biased.
The trial judge had no pecuniary or personal interest in the outcome of this case, she had
not been the target of personal abuse or criticism from defendant, she had not been
enmeshed in other matters involving defendant, and she had not previously participated as
an accuser, investigator, fact-finder, or initial decision-maker in defendant’s case. See
Cain, 451 Mich at 498.
2. RIGHT TO BE PRESENT AND RIGHT TO COUNSEL
Even if the trial judge’s ex parte communications with Prosecutor Koch do not show
that she was actually biased, defendant argues, the trial judge’s ex parte communications
deprived him of his constitutional right to be present and his right to counsel. Again, we
disagree.
The Due Process Clauses and the Confrontation Clauses of the Michigan
Constitution and the United States Constitution impliedly guarantee defendants the right to
be present at any stage of a criminal proceeding in which their substantial rights might be
adversely affected. See People v Mallory, 421 Mich 229, 247; 365 NW2d 673 (1984)
(noting that a defendant has the right to be present during any stage of a criminal proceeding
in which his substantial rights might be affected); Gagnon, 470 US at 526-527 (noting that
a defendant’s presence is required when necessary to ensure fundamental fairness or a
24
reasonably substantial opportunity to defend against the charge). In addition, the Sixth
Amendment and Const 1963, art 1, § 20, guarantee to the accused the right to counsel at
all critical stages of a criminal proceeding, see Montejo v Louisiana, 556 US 778, 786; 129
S Ct 2079; 173 L Ed 2d 955 (2009); People v Ackley, 497 Mich 381, 388; 870 NW2d 858(2015), which has likewise been interpreted to mean that a defendant has the right to
counsel at any stage at which a defendant’s substantial rights may be affected, see Mempa
v Rhay, 389 US 128, 134; 88 S Ct 254; 19 L Ed 2d 336 (1967); Coleman v Alabama, 399
US 1, 9; 90 S Ct 1999; 26 L Ed 2d 387 (1970); People v Hammerquist, 107 Mich App 771,
777; 309 NW2d 637 (1981); see also People v Buie, 298 Mich App 50, 61; 825 NW2d 361(2012) (holding that the right to counsel extends to all critical stages of the proceedings
where counsel’s absence might harm the defendant’s right to a fair trial). Accordingly, to
determine whether the trial judge and Prosecutor Koch’s ex parte exchange deprived
defendant of his right to be present or to counsel, the salient question is whether the
interaction was one in which defendant’s substantial rights could have been affected.
It was not. During the interaction between the trial judge and Prosecutor Koch, the
trial judge was not considering whether to take some action that would affect defendant,
and Prosecutor Koch was not suggesting or implying that the trial judge should take some
action that would do so. In response to witness testimony, the trial judge e-mailed
Prosecutor Koch expressing concern about law enforcement’s missteps and asking why
those missteps occurred. Prosecutor Koch answered explaining why the missteps occurred
and how law enforcement and the prosecutor’s office would prevent such mistakes moving
forward. This is not a case in which the prosecution was privately advocating to the judge
about what sentence defendant should receive, see United States v Wolfson, 634 F2d 1217,
25
1221 (CA 9, 1980); in which the judge was trying to decide whether to admit evidence at
the defendant’s trial, see Yohn v Love, 76 F3d 508, 516-517, 522 (CA 3, 1996); or in which
the prosecution and the judge were deliberating privately about what to do about a
potentially partial juror, see Barnwell, 477 F3d at 847-849, 851-853. This is a case in
which the trial judge and Prosecutor Koch had a brief, albeit inappropriate, side
conversation about law enforcement’s negligence, a topic of which defendant and defense
counsel were fully aware. 16 Simply put, the brief e-mail exchange between the trial judge
and Prosecutor Koch was not a critical stage of the proceedings. 17
Even if the trial judge’s ex parte interaction with Prosecutor Koch was not a critical
stage of the proceedings, defendant contends, the ex parte interaction still deprived him of
his due-process right to a fair trial. This is because, according to defendant, the trial
prosecutor altered her strategy in response to the trial judge’s comments to Prosecutor
Koch.
16
Defendant suggests that he was deprived of counsel or the right to effective assistance
of counsel because, had the trial judge disclosed her ex parte communications with
Prosecutor Koch to defense counsel, defense counsel could have moved for the trial judge’s
disqualification and for a mistrial. Even if defense counsel could have done so, this
presumes that defendant had a constitutional right to the presence of counsel during the
trial judge’s exchange in the first place, which, as discussed, he did not. Although Canon
3(C) might have ethically obligated the trial judge to disclose her exchange with Prosecutor
Koch and to raise the issue of her disqualification sua sponte, nothing in our state or federal
Constitution obligated her to do so.
17
We of course acknowledge that defendant’s trial is a critical stage of the proceedings and
that the ex parte exchange here occurred during this critical stage. Our point is that the
brief e-mail exchange was not itself a critical stage, given that during this exchange, the
trial judge was not considering whether to take some action that would affect defendant,
and Prosecutor Koch was not suggesting or implying that the trial judge should take some
action that would do so.
26
To begin, we reiterate that this is not an instance in which the trial judge explicitly
flagged for correction perceived issues in the prosecution’s case or else offered the
prosecution advice about trial strategy. Again, the trial judge sent a few brief e-mails
expressing her concern about the sufficiency of law enforcement’s investigation and why
missteps occurred. She did not say anything to suggest that she was concerned about the
trial prosecutor’s ability to prove defendant’s guilt or about the trial prosecutor’s trial
strategy, nor did she offer advice about trial strategy. 18 It remains undisputed that the jury
was unaware of the trial judge’s communications with Prosecutor Koch, and there is no
evidence that these communications affected how the jury was instructed or the substance
of the jury’s deliberation over a verdict.
Defendant does not contend otherwise. Rather, his theory seems to be that if the
trial judge’s ex parte communications led the trial prosecutor to alter her strategy, then this
alone is enough to establish a due-process violation, implying that it does not matter
whether the trial judge intended to induce the trial prosecutor to do so. Assuming without
deciding that defendant’s theory is correct, we see nothing showing that the trial prosecutor
altered her strategy in response to the trial judge’s exchange with Prosecutor Koch. To
start, there is nothing showing that Prosecutor Koch told the trial prosecutor about the trial
judge’s comments. And even if we were to impute Prosecutor Koch’s knowledge to the
trial prosecutor, we see nothing to suggest that the trial prosecutor tried a different tack in
response to the trial judge’s comments. The trial prosecutor disclosed the trooper’s
investigative missteps in her opening statement, solicited testimony from the trooper about
18
We therefore disagree with Justice BERNSTEIN’s claim that the trial judge “indicat[ed]
what she believed to be deficiencies in the prosecution’s case-in-chief.” Post at 6.
27
his missteps, and solicited testimony from Detective Workman about law enforcement’s
failure to have the victim undergo a medical examination—all before the trial prosecutor
could have known about the trial judge’s ex parte communications. Indeed, the trial judge
e-mailed Prosecutor Koch while the trooper was already testifying and while Detective
Workman was already testifying. And the trial prosecutor stayed consistent after the trial
judge’s ex parte communications: the trial prosecutor mentioned the trooper’s investigative
missteps again in her closing statement. Though she did not mention law enforcement’s
failure to have the victim undergo a medical examination, neither did she do so in her
opening statement. Defendant suggests that the trial prosecutor spent more time in her
closing statement emphasizing the trooper’s shortcomings than she did in her opening
statement, and that she mentioned the trooper’s investigative shortcomings in her opening
statement only to explain why the jury would not see any DNA evidence. Yet this is to be
expected. Opening statements are an attorney’s opportunity to give the jury an overview
of the evidence, not to make arguments about the defendant’s guilt or innocence. All said,
we see nothing to suggest that the trial prosecutor altered her strategy in response to the
trial judge’s ex parte communications with Prosecutor Koch. Defendant has therefore
failed to show that the trial judge’s ex parte communications violated his due-process right
to a fair trial on this basis.
III. CONCLUSION
As we have made clear, we believe that the trial judge’s conduct in this case violated
the Michigan Code of Judicial Conduct. The trial judge’s actions fell short of the high
ethical standards that Michigan jurists are expected to uphold, and regrettably, her behavior
28
has the potential to erode public confidence in the integrity of our justice system. Still, the
issue before us is not whether the trial judge should be sanctioned for her misconduct. The
issue before us is whether the trial judge’s violation of a judicial canon provided the trial
court a legal basis on which to grant defendant a new trial. And because the trial judge’s
failure to recuse herself did not result in a miscarriage of justice at defendant’s trial or
deprive defendant of any constitutional right, we conclude that the trial court had no such
legal basis. Accordingly, the Court of Appeals correctly concluded that the trial court
abused its discretion by granting defendant a new trial under MCR 6.431(B). The judgment
of the Court of Appeals is affirmed.
Elizabeth T. Clement
Brian K. Zahra
David F. Viviano
29
STATE OF MICHIGAN
SUPREME COURT
PEOPLE OF THE STATE OF MICHIGAN,
Plaintiff-Appellee,
v No. 164133
DANIEL ALBERT LOEW,
Defendant-Appellant.
BOLDEN, J. (concurring in part, dissenting in part, and concurring in the judgment).
I agree with the lead opinion that defendant has not shown it to be a miscarriage of
justice that the trial judge declined to recuse herself after having engaged in ex parte
communications with Allegan County Prosecutor Myrene Koch. 1 I further agree that this
Court should not adopt the test the United States Supreme Court outlined in Liljeberg v
Health Servs Acquisition Corp, 486 US 847; 108 S Ct 2194; 100 L Ed 2d 855 (1988), as
defendant and his supportive amici ask us to do. Therefore, I join the lead opinion in
affirming the Court of Appeals. However, I depart from the lead opinion because I would
not conclude through these proceedings that the trial judge violated the Michigan Code of
Judicial Conduct. I write separately to explain why I would have arrived at different
conclusions from the lead opinion on this issue and also to explain my reasons for
1
To align with the terminology used in the lead opinion, I will refer to Judge Margaret
Bakker as “the trial judge,” Myrene Koch as “Prosecutor Koch,” and the prosecuting
attorney who led the trial in which defendant was convicted as “the trial prosecutor”
throughout my opinion.
agreement that the Liljeberg test should not be adopted. For the reasons established below,
I concur in part, but in part, I respectfully dissent.
I. ANALYSIS
A. IT IS UNNECESSARY TO DETERMINE WHETHER THE CODE OF JUDICIAL
CONDUCT WAS VIOLATED, NOR SHOULD THIS COURT ADDRESS THAT
QUESTION IN THESE PROCEEDINGS
The lead opinion concludes that the trial judge’s exchange with the elected
prosecutor was unethical and in violation of the Michigan Code of Judicial Conduct. But
whether the trial judge has violated the judicial canons is not part of this appeal, and this
Court should not and need not make that determination in this case. 2 Importantly,
2
The Judicial Tenure Commission was established by the Michigan Constitution and is
governed by our court rules. Const 1963, art 6, § 30; MCR 9.200 et seq. Subchapter 9.200
of the Michigan Court Rules outlines the way the Commission investigates allegations of
judicial misconduct, what constitutes a violation, the procedures to be followed during and
after an investigation, possible defenses, and the rights and responsibilities of all
individuals involved. Opining on the merits of a judicial-conduct violation at this juncture
is premature. Moreover, doing so would bypass the very rules designed to ensure that any
conclusion that the judicial canons were violated is reached—and any possible
ramifications of that decision ensue—only according to the procedures outlined in our
court rules.
Although the lead opinion implies that its holding does not implicate the role of the
Commission because it gives no commentary on whether the trial judge should be
sanctioned for this conduct, its conclusion that a violation of the judicial canons occurred
nevertheless significantly undermines the role of the Commission in investigating claims
and determining whether conduct constitutes a violation. Our court rules have expressly
assigned significant investigatory powers to the Commission. For example, they may take
evidence, require an investigated judge to comment on the investigation, and issue
subpoenas to collect information. MCR 9.221. Additionally, it is not until the Commission
determines there is sufficient evidence to believe a judge has engaged in misconduct that
the Commission may issue a formal complaint against the judge. MCR 9.224. Given that
this Court has superintending control over the Commission, it is difficult to imagine how
any decisions we make that the judicial canons were violated would not in some way affect
the investigatory process we expect the Commission to follow for every judge whose
2
discussing whether the trial judge violated the judicial canons shifts the focus from the crux
of the issue in this case—whether defendant had a fair trial.
This Court has not yet addressed ex parte communications between a trial judge and
a county’s lead prosecutor, but it has addressed ex parte communications between a judge
and a deliberating jury. People v France, 436 Mich 138; 461 NW2d 621 (1990). In
France, this Court classified the ex parte communications between the judge and jury into
three categories: substantive, administrative, and housekeeping. Id. at 142-143.
“Substantive communication encompasses supplemental instructions on the law given by
the trial court to a deliberating jury.” Id. at 143. A substantive communication “carries a
presumption of prejudice in favor of the aggrieved party,” and “[t]he presumption may only
be rebutted by a firm and definite showing of an absence of prejudice.” Id. 3 I agree with
conduct is reviewed by the Commission. This Court has previously declined to address
allegations of judicial misconduct that were neither found nor recommended for discipline
by the Commission, and I cannot discern a meaningful reason why we would do so now.
See, e.g., In re Simpson, 500 Mich 533, 569-570; 902 NW2d 383 (2017) (“Another
compelling reason to limit our review in [Commission] proceedings to allegations of
misconduct found and recommended to us by the [Commission] is that a respondent judge
is entitled to notice of the charges and a reasonable opportunity to respond to them.
Without such notice, it is not clear to us how a respondent judge would know which charges
are at issue and, therefore, which ones he or she should substantively address when a case
proceeds to our Court. Is our review limited to the charges in the formal complaint or an
amended version of it? Or the findings of the master? Or the findings and
recommendations of the [Commission]? Should a respondent and his or her attorney be
put in the untenable position of having to argue against possible findings of misconduct
that were not charged in the complaint or made by either the master or the [Commission]
but might be discerned by a member of this Court? Whatever could be said about such a
regime, we would no longer say that it ‘provides a full panoply of procedural guarantees
for adjudicating allegations of judicial misconduct.’ ”).
3
France is not on point because these categories refer to the types of communications that
can take place between the trial judge and a deliberating jury. France explains that an
3
the lead opinion that the communication in the instant case was a substantive
communication and that the Court of Appeals erred by categorizing this as an
administrative communication.
Although France was decided in a different context, it is instructive. This Court
analyzed the judicial ex parte communications to determine whether the defendant had a
fair trial. Id. at 166. It concluded that the judicial ex parte communications at issue there
did not deprive the defendant of a fair trial because they involved administrative
communications, not substantive communications. Id. However, at no point did this Court
weigh in on whether the judicial canons were violated. Violations of the judicial code were
not at issue in France, and they should not be at issue here. When we consider whether a
defendant’s trial was fair, the focus of our analysis ought to be limited to just that—whether
defendant’s trial was fair.
B. LILJEBERG IS AN INADEQUATE LEGAL SOLUTION TO THIS PROBLEM
Defendant argues that he is entitled to a new trial because we should adopt the test
articulated in Liljeberg, which would compel that result. However, Liljeberg is not
administrative communication would involve, for example, the availability of evidence for
deliberations or instructions about when to continue deliberating, whereas housekeeping
communications would be the equivalent of receiving lunch orders from the jurors. Id. at
143. However, France does not adequately resolve the question at issue because ex parte
communications with a jury are categorically different than ex parte communications with
a lead prosecutor. And beyond that, it would be almost impossible to apply the France test
to ex parte communications with the lead prosecutor. For example, it would seem that the
entire category of “housekeeping” communications would be rendered nugatory if France
were applied to conversations with the prosecution. Although France is an inadequate test
here, it is helpful in showing that this Court treats more seriously communications that are
more closely related to the fair-trial right.
4
binding, serves to remedy a problem in a different context, and would provide an
inadequate solution to the issue before us.
1. LILJEBERG AND THE APPEARANCE OF IMPROPRIETY
Liljeberg began as a federal civil lawsuit concerning declaratory relief. The
defendant, John Liljeberg, Jr., had formed a corporation called St. Jude Hospital of Kenner,
Louisiana, in order to negotiate a purchase of land from Loyola University that could be
rezoned and used for the building of a hospital in Kenner. Liljeberg, 486 US at 850-852.
The Loyola University Board of Trustees received regular reports on the progress of these
negotiations because if a deal was reached, Loyola University stood to financially benefit
from both the sale and the related rezoning of the adjacent property. Id. at 853. Meanwhile,
the defendant reached a preliminary agreement with an entity then known as Hospital
Affiliates International (HAI), which was to purchase a second tract of land for the hospital
and retain the defendant as a hospital consultant in exchange for the transfer of ownership
of St. Jude from the defendant to HAI. Id. at 853-854.
On November 30, 1981, HAI’s corporate successor, Health Services Acquisition
Corporation, brought an action against the defendant, seeking a declaratory judgment that
the preliminary agreement had transferred ownership of St. Jude to HAI without the need
for a separate finalizing agreement. Id. at 854-855. The case was decided in the
defendant’s favor after a two-day bench trial conducted by Judge Robert Collins, who
announced his intended ruling from the bench at the close of evidence on January 22, 1982,
and entered a judgment in favor of the defendant on March 16, 1982. Id. at 855-857.
5
About 10 months after the judgment entered, Health Services Acquisition
Corporation learned that Judge Collins was a member of the Loyola University Board of
Trustees. Id. at 850. Notably, Judge Collins had attended a board meeting on
November 12, 1981, at which the trustees discussed the ongoing negotiations, providing
Judge Collins with actual knowledge of Loyola University’s potential interest with St.
Jude. Id. at 856. Eventually, and while Judge Collins was overseeing the case, Loyola
University formally agreed to sell the property to the defendant on March 19, 1982, after
discussing the matter at a board meeting on January 28, 1982. Id. at 857. Judge Collins
did not attend that meeting, but the minutes were mailed to him on March 12, 1982, and
delivered on March 14 or March 15, shortly before he entered a judgment in the defendant’s
favor on March 16. Judge Collins claimed that he did not open them until March 24, 1982,
in preparation for a board meeting the next day. Id. at 855, 858.
Health Services Acquisition Corporation moved for relief from the judgment,
arguing that the judge was disqualified under 28 USC 455 at the time the judgment was
entered. 4 Id. at 850. Judge Collins denied the motion. Id. at 851. On appeal, the United
4
28 USC 455 is the federal statute concerning disqualification of a federal justice, judge,
or magistrate. The portions of the statute relevant to Liljeberg provide:
(a) Any justice, judge, or magistrate judge of the United States shall
disqualify himself in any proceeding in which his impartiality might
reasonably be questioned.
(b) He shall also disqualify himself in the following circumstances:
* * *
(4) He knows that he, individually or as a fiduciary, or his spouse or
minor child residing in his household, has a financial interest in the subject
6
States Court of Appeals for the Fifth Circuit vacated the judgment because “ ‘Judge Collins
should have recused himself when he obtained actual knowledge” of Loyola’s potential
interest with St. Jude. Id. at 858 (citation omitted).
The United States Supreme Court granted certiorari to determine whether Judge
Collins had violated 28 USC 455 and, if so, the proper remedy. Ultimately, the Supreme
Court affirmed the judgment of the Fifth Circuit. Noting that the purpose of the statute
was to promote public confidence in judicial integrity, the Court held that a judge should
recuse themselves if they have actual knowledge of an impropriety or concludes that their
“impartiality might reasonably be questioned.” Id. at 860-861. The Supreme Court
concluded that Judge Collins had violated 28 USC 455 by failing to disqualify himself after
he became aware of an appearance of impropriety. Id. at 861. However, the Supreme
Court noted that a simple violation would not necessarily trigger a new trial. Id. at 862.
Instead, the Supreme Court looked critically at the facts that might reasonably have caused
an objective observer to question Judge Collins’ impartiality. Id. at 865.
In determining whether a valid judgment should be vacated for a violation of 28
USC 455, the Supreme Court held that “it is appropriate to consider the risk of injustice to
the parties in the particular case, the risk that the denial of relief will produce injustice in
matter in controversy or in a party to the proceeding, or any other interest
that could be substantially affected by the outcome of the proceeding.
* * *
(c) A judge should inform himself about his personal and fiduciary
financial interests, and make a reasonable effort to inform himself about the
personal financial interests of his spouse and minor children residing in his
household.
7
other cases, and the risk of undermining the public’s confidence in the judicial process.”
Id. at 864. Applying this test, the Supreme Court found four facts particularly alarming:
(1) the negotiations were the subject of discussion by the trustees at a meeting attended by
Judge Collins days before the lawsuit was filed, and it was “remarkable that the
judge . . . completely forgot about [Loyola’s] interest in having a hospital constructed on
its property in Kenner”; (2) had Judge Collins opened the meeting minutes that he received
on March 14 or March 15, he would have had an obligation to recuse himself before
entering the final judgment; (3) Judge Collins failed to recuse himself on March 24 when
he reviewed the meeting minutes and knew that he, either in his own capacity or as a
fiduciary, had a financial interest in the subject matter in controversy, which the Supreme
Court found “quite inexcusable”; and (4) Judge Collins did not acknowledge that he knew
about Loyola’s interest or that he had a duty to recuse himself when Health Services
Acquisition Corporation moved to vacate the judgment. Id. at 865-867. The Court stated
that these facts “create[d] precisely the kind of impropriety that [28 USC 455] was intended
to prevent,” noted that providing relief in the case might prevent substantive injustice in
future cases by encouraging judges and litigants to more carefully examine the possible
grounds for disqualification and disclosure, and concluded that there was a greater risk of
unfairness in upholding the judgment than in allowing a new judge to reconsider the case.
Id. at 868. The Fifth Circuit’s judgment ordering vacating of the judgment and a new trial
was affirmed. Id. at 869.
8
2. LILJEBERG DOES NOT WORK IN THIS CONTEXT
Defendant asks us to adopt Liljeberg in this context and require that the trial judge’s
ex parte communications with the county’s lead prosecutor be examined in this light. In
other words, defendant wants us to analyze and weigh the following three considerations
to determine whether a new trial is appropriate: (1) the risk of injustice to the parties in this
case, (2) the risk that denying relief will produce injustice in other cases, and (3) the risk
of undermining the public’s trust in the judiciary. On the surface, adopting such an
approach has some appeal because it would allow this Court to anchor a trial judge’s ex
parte communications to the larger question of how the public would perceive those
communications. However, I disagree with this approach. Liljeberg does not work in this
context.
First, it must be said that Liljeberg is not binding on this decision. In that case, the
threshold question that the Supreme Court reached was whether 28 USC 455—the statute
concerning disqualification of federal judges—had been violated. That statute does not
pertain to this case, which involves a state judge, not a federal judge. There is no parallel
statute at issue here. 5 Second, Liljeberg’s holding was rooted in the context of a trial
judge’s failure to disclose a financial arrangement that may have appeared to have placed
the judge in a self-interested position while overseeing a civil trial concerning declaratory
5
Although the existence of a statutory scheme might not make a difference to whether
relief is warranted in a given case, Liljeberg presented a question of statutory interpretation,
and the Supreme Court relied heavily on legislative-purpose statements in crafting its test
and remedy. See, e.g., Liljeberg, 486 US at 849, 859-860. That there is no parallel statute
in Michigan does not negate the many reasons I describe throughout this opinion for why
relief may be available in this context; it merely signifies one more reason why Liljeberg
is neither binding nor persuasively applied in this context.
9
relief. The instant case concerns a criminal trial, in which the trial judge was not alleged
to have any financial interest. Instead, this case involves ex parte communications initiated
and continued by the trial judge to the county prosecutor about defendant’s case during
defendant’s criminal trial. The rights at stake, the ongoing nature of the communications
during and about defendant’s trial, and the gravity of the potential consequences make it
particularly difficult to modify Liljeberg’s test for application in this and other criminal
cases.
Courts that have tried to apply Liljeberg in criminal contexts have inadequately
balanced the rights and interests at stake. The United States Court of Appeals for the
Seventh Circuit is an example, as demonstrated by United States v Atwood, 941 F3d 883
(CA 7, 2019); United States v Williams, 949 F3d 1056 (CA 7, 2020); and United States v
Orr, 969 F3d 732 (CA 7, 2020). All three cases were appeals of federal convictions and
concerned Judge Colin Bruce, a federal trial judge in Illinois who was investigated by the
Seventh Circuit Judicial Council and suspended from the bench for one year for engaging
in ex parte communications with the United States Attorney’s Office for the Central
District of Illinois, where he had worked for 24 years before assuming his role on the bench.
See Shannon v United States, 39 F4th 868, 883 (2022), citing In re Complaints Against
Dist Judge Colin S Bruce, order of the Judicial Council of the Seventh Circuit, entered May
14, 2019 (Complaint Nos. 07-18-90053 and 07-18-90067). Judge Bruce had presided over
the defendants’ criminal trials in Atwood, Williams, and Orr. A different judge conducted
the sentencing hearings in Williams and Orr. The Seventh Circuit applied Liljeberg in each
case, on the basis of the defendants’ assertions that 28 USC 455 had been violated, to
determine whether relief was available.
10
The first of these three cases was Atwood. There, the defendant was convicted and
sentenced to serve 210 months in federal prison for federal drug crimes. Atwood, 941 F3d
at 883-884. While his case was pending, Judge Bruce engaged in ex parte communications
with the United States Attorney’s Office about other cases. Id. The Seventh Circuit applied
Liljeberg to hold that defendant was entitled to resentencing before a new judge. Id. The
court reasoned that the first factor—the risk of injustice in the case at bar—favored
resentencing. The court explained that the open-ended sentencing factors leave room for
the judge to exercise discretion, creating a potential unfairness to the defendant, whereas
there would be little risk of unfairness to the government beyond its incurring some
administrative costs. Id. at 885. The second factor, the risk of injustice to future litigants,
favored resentencing because doing so would encourage judges to “exercise caution in their
communications.” Id. The third factor, the risk of undermining public confidence in the
impartiality of the judiciary, also favored resentencing, given the uncontested fact that
Judge Bruce had “compromised his appearance of impartiality.” Id. at 886.
Next, the Seventh Circuit decided Williams. There, the defendant had been jury-
convicted of obstruction of commerce by robbery in a trial over which Judge Bruce
presided. Williams, 949 F3d at 1058. Judge Bruce did not preside over the defendant’s
sentencing. Id. Among other things, the defendant appealed his conviction on the basis of
Judge Bruce’s violation of 28 USC 455. Id. The Seventh Circuit applied Liljeberg and
concluded that no new trial was warranted. Id. at 1066. The first factor did not favor a
new trial because none of the ex parte communications concerned the defendant’s case.
The court explained that the conviction was reached by a jury, Judge Bruce made minimal
rulings before and during trial, and “[t]he government would likely spend valuable time
11
and money to retry this case thereby diverting resources from other cases.” Id. at 1065.
The second factor did not favor a new trial because Judge Bruce had already been publicly
reprimanded and changed his communication policies, so there was no risk of prejudice in
future cases if relief was denied. Id. Nor did the third factor favor a new trial, because
Judge Bruce had not undermined the public’s confidence in the judicial process given that
the trial was by jury and his discretionary rulings had no significant effect on the outcome.
Id.
The third case was Orr. There, as in Williams, Judge Bruce presided over the
defendant’s jury trial but was reprimanded before the defendant was sentenced. Orr, 969
F3d at 734-736. As in Williams, the defendant sought a new trial on the basis of Judge
Bruce’s violation of 28 USC 455. The Seventh Circuit applied the Liljeberg factors to hold
that Judge Bruce’s decision to not disqualify himself from the case was not harmless, and
it vacated the defendant’s conviction. Id. at 742. The first factor favored a new trial
because Judge Bruce had made discretionary decisions at trial and it was possible that
personal biases influenced the outcome of those decisions and, thus, the case. Id. at 741.
The court concluded that the risk of unfairness to the government was slight because the
prosecution’s case was brief and straightforward: the defendant faced one charge and the
trial lasted only two days. Id. The second factor did not favor a new trial for the same
reasons asserted in Williams. Id. The third factor favored a new trial because, although
the case was decided by a jury, Judge Bruce made discretionary decisions to admit evidence
of the defendant’s drug dealing and to permit the prosecution to cross-examine the
defendant on a felony conviction for drug dealing. Id. at 741-742. Because these
discretionary decisions “bolstered the prosecution’s case, which rested on circumstantial
12
evidence and credibility calls,” upholding the conviction could have “damage[d] the
public’s confidence in the impartiality of the judiciary.” Id. at 742.
Examining the ways in which the Seventh Circuit applied Liljeberg in criminal
contexts demonstrates why it provides an inadequate legal analysis or remedy for the issue
currently before us. 6 Given that the central issue in this case pertains to whether defendant
is entitled to a new trial because of the presiding judge’s ex parte communications during
his trial, Williams and Orr provide the best points of comparison. In Williams, the
defendant was not entitled to a new trial, but in Orr, the defendant was. The difference
between the two hinged on two considerations: (1) whether the trial judge had exercised
their discretion over evidentiary matters, and (2) the length or complexity of the
defendant’s trial.
Because I would conclude that the crux of this issue is whether defendant’s trial was
fair and that this question is best resolved by constitutional analysis, Liljeberg comes up
short. Whether a defendant’s trial is lengthy or complicated and whether the presiding
judge exercised their discretion over evidentiary matters are hardly metrics through which
one’s constitutional rights should be measured. Adopting Liljeberg in this context would
require us to adopt a constitutional test under which an individual with a longer, more
complicated trial is less likely to be awarded a new trial for the same constitutional error—
not on the basis of the evidence presented against them, but solely because it was so
6
The Seventh Circuit had a stronger rationale for applying Liljeberg in the Judge Bruce
cases than we would have in this context given that the defendants in those cases all
requested a new trial or resentencing hearing on the basis of a violation of 28 USC 455.
No such violation is at issue here, given that 28 USC 455 pertains to disqualification of
federal judges.
13
expensive to try them in the first place. Constitutional rights are not measured through this
lens. Efficiency is not the barometer for constitutional rights. See People v Jemison, 505
Mich 352, 364; 952 NW2d 394 (2020) (explaining that, when considering whether the
defendant’s Sixth Amendment right to confront an adverse witness was violated, “expense
is not a justification for a constitutional shortcut”).
The analyses from Williams and Orr demonstrate a problem with trying to translate
Liljeberg from its original civil context into the criminal context. The root of this problem
lies in its first prong, which requires a reviewing court to balance the risk of unfairness to
both parties, that is, the defendant and the prosecution in a criminal case. The Sixth
Amendment of the United States Constitution articulates rights held by criminal
defendants. It explains:
In all criminal prosecutions, the accused shall enjoy the right to a
speedy and public trial, by an impartial jury of the State and district wherein
the crime shall have been committed, which district shall have been
previously ascertained by law, and to be informed of the nature and cause of
the accusation; to be confronted with the witnesses against him; to have
compulsory process for obtaining witnesses in his favor, and to have the
Assistance of Counsel for his defence. [US Const, Am VI].[7]
When the Constitution singles out rights to be enjoyed by “the accused,” it is making clear
that the rights pertaining to a fair trial are rights afforded to criminal defendants, not to all
involved parties. The fair-trial rights are not about balancing these interests with the
fairness to the prosecution. Any test that purports to sacrifice a criminal defendant’s
constitutional rights by balancing them against the risk of unfairness to the prosecution will
7
Michigan’s Constitution has a similar provision articulating the rights of criminal
defendants. See Const 1963, art 1, § 20.
14
ultimately provide inadequate protection of those rights. When Liljeberg has been
extended to criminal cases, courts have imposed this balancing test and awarded new
trials—but, seemingly, only after concluding that it was efficient to do so. Thus, Liljeberg
provides an inadequate solution to this legal issue. 8
II. CONCLUSION
I consider the issue before us to be of utmost importance to the Court. Criminal
defendants are afforded constitutional guarantees of fair trials and due process of law.
Protecting those rights is one of the judiciary’s most important roles. Here, when the trial
judge initiated and continued ex parte conversations with Prosecutor Koch—about matters
related to defendant’s trial while the judge was overseeing it—the trial judge may have
failed to protect those rights. However, whether the trial judge violated the judicial canons,
in my view, is unnecessary for this Court to address while determining whether there was
either a constitutional violation or a miscarriage of justice.
Although defendant seeks to adopt Liljeberg as the framework for resolving these
issues, I agree with the lead opinion that Liljeberg does not provide the proper approach
for criminal trials. Thus, defendant has not demonstrated entitlement to relief, and I join
the lead opinion in affirming the judgment of the Court of Appeals.
Kyra H. Bolden
8
Having agreed with the lead opinion that Liljeberg is inapplicable, the next question ought
to be which test we should apply. I agree with the lead opinion and Justice BERNSTEIN that
the test employed by United States v Barnwell, 477 F3d 844 (CA 6, 2007), is the closest
applicable test. However, defendant almost exclusively asks us to adopt Liljeberg, and his
brief references to Barnwell do not lead me to question the lead opinion’s analysis of that
case under these facts. Thus, I agree with the lead opinion’s decision to affirm the Court
of Appeals.
15
STATE OF MICHIGAN
SUPREME COURT
PEOPLE OF THE STATE OF MICHIGAN,
Plaintiff-Appellee,
v No. 164133
DANIEL ALBERT LOEW,
Defendant-Appellant.
WELCH, J. (concurring in part and dissenting in part).
A judge who engages in unethical behavior is subject to investigation by the Judicial
Tenure Commission and disciplinary action from this Court. A judge’s unethical behavior
on the bench can also undermine the integrity of and public confidence in not only an
individual trial, but also the judiciary as a whole. Unethical judicial behavior that taints
the integrity of a criminal proceeding may require retrying or resentencing the accused. On
this, the lead opinion and I are aligned. I concur in the lead opinion’s conclusion that the
ex parte e-mail communications between Allegan County Circuit Court Judge Margaret
Bakker and the elected Allegan County prosecutor, Myrene Koch, during defendant’s
criminal trial were unethical and violative of the Michigan Code of Judicial Conduct and
MCR 2.003(C)(1)(b)(ii) such that recusal was required. I also agree with the lead opinion
that, under the facts of this case, the ex parte communications did not violate defendant’s
constitutional rights to due process, to be present during critical stages of a proceeding, or
to counsel, and thus there was not a constitutional error that would mandate granting
defendant a new trial.
Despite the lack of a constitutional violation, I conclude that the Court is not
precluded from holding that a miscarriage of justice occurred for purposes of MCL 769.26
and MCR 6.431(B). I would not apply traditional harmless error review under these
circumstances and would instead augment the harmless error analysis by adopting the test
articulated in Liljeberg v Health Servs Acquisition Corp, 486 US 847; 108 S Ct 2194; 100
L Ed 2d 855 (1988). The Liljeberg test was crafted precisely for situations where judicial
misconduct creates an appearance of impropriety that undermines the integrity of a judicial
proceeding but does not rise to the level of a constitutional violation. As a result of
adopting a new standard, I would remand this matter to a trial court judge other than Judge
Bakker to apply the Liljeberg test. To the extent the lead opinion takes a different approach
and reaches a different conclusion, I respectfully dissent.
I. FACTUAL BACKGROUND
Defendant was charged with multiple counts of criminal sexual conduct (CSC)
arising from allegations that he had repeatedly sexually assaulted a minor cousin of
defendant’s then fiancée. As recounted at length by Chief Justice CLEMENT, in the middle
of defendant’s trial, Judge Bakker initiated and exchanged several e-mail communications
with Prosecutor Koch. The e-mails discussed the collection of evidence during the
underlying police investigation and the failure to refer the complainant for a medical
examination following her disclosure of the allegations against defendant. Neither Judge
Bakker nor Prosecutor Koch disclosed these ex parte communications to defendant or his
attorney. The jury found defendant guilty of several counts of CSC-I, CSC-II, and CSC-
III. Defendant was later sentenced by Judge Bakker to 240 to 480 months’ incarceration
2
for the CSC-I counts and 240 to 360 months’ incarceration for the CSC-II and CSC-III
counts.
Defendant appealed by right to the Court of Appeals. While that appeal was pending
and nearly a year after his conviction, defendant learned of the ex parte communications
between Judge Bakker and Prosecutor Koch. The ex parte communications were revealed
because of a Freedom of Information Act request that was filed with the county
prosecutor’s office by an unrelated third party. Defendant then filed a motion requesting
an evidentiary hearing and a new trial under MCR 6.431(A)(2) and MCR 7.208(B)(1),
alleging that he had been deprived of the right to a fair trial as a result of judicial misconduct
and bias, ineffective trial counsel, and prosecutorial misconduct. After Circuit Court
Judges Bakker and Robert Kengis recused themselves, and following a hearing, District
Court Judge William Baillargeon granted defendant’s motion.
Defendant and the prosecution each sought relief in the Court of Appeals. In a split,
published decision, a majority of the Court of Appeals reversed the order granting
defendant a new trial for the reasons aptly set forth in Chief Justice CLEMENT’s opinion.
Judge RIORDAN filed a dissenting opinion. Defendant sought leave to appeal in this Court,
which was granted. People v Loew, 510 Mich 952 (2022).
II. AN APPEARANCE OF IMPROPRIETY AND A MISCARRIAGE OF JUSTICE
Canon 3(A)(4) of the Michigan Code of Judicial Conduct prohibits a judge from
communicating with a party to a legal proceeding outside the presence of opposing counsel
in most instances. Canon 3(A)(4) provides, in part:
A judge shall not initiate, permit, or consider ex parte
communications, or consider other communications made to the judge
3
outside the presence of the parties concerning a pending or impending
proceeding, except as follows:
(a) A judge may allow ex parte communications for scheduling,
administrative purposes, or emergencies that do not deal with substantive
matters or issues on the merits, provided:
(i) the judge reasonably believes that no party or counsel for a party
will gain a procedural or tactical advantage as a result of the ex parte
communication, and
(ii) the judge makes provision promptly to notify all other parties and
counsel for parties of the substance of the ex parte communication and allows
an opportunity to respond.
Conduct that creates an appearance of impropriety is proscribed by Canon 2(A), which
provides as follows:
A Judge Should Avoid Impropriety and the Appearance of Impropriety in
All Activities.
A. Public confidence in the judiciary is eroded by irresponsible or
improper conduct by judges. A judge must avoid all impropriety and
appearance of impropriety. A judge must expect to be the subject of constant
public scrutiny. A judge must therefore accept restrictions on conduct that
might be viewed as burdensome by the ordinary citizen and should do so
freely and willingly. [Emphasis added.]
A. THE JUDGE’S VIOLATION OF CANONS 3(A)(4) AND 2(A)
REQUIRED RECUSAL
Like the lead opinion, I conclude that the communications Judge Bakker initiated
with Prosecutor Koch were substantive and of such a nature that a reasonable and objective
person would question the judge’s impartiality. Judge Bakker reached out to Prosecutor
Koch during trial while evidence was being presented by an assistant prosecutor working
under Koch. Specifically, Judge Bakker raised concerns with aspects of the underlying
police investigation and the evidence being presented.
4
As the lead opinion observes, “[n]ot only did the trial judge give Prosecutor Koch
‘private access to’ her ear, considering the contents of her communications, one might
reasonably question whether the trial judge was interested in seeing the prosecution
succeed or seeing defendant convicted.” Ante at 14. Even in the absence of direct
evidence, one might reasonably wonder whether the substance of the judge’s
communications and concerns were conveyed to the assistant prosecutor who was trying
defendant’s case and, if so, might this have shaped the prosecution’s strategy for closing
out the trial. These concerns are heightened when one considers Judge Bakker’s and
Prosecutor Koch’s long history of working together as prosecutors in Allegan County
before assuming their current roles.
As Court of Appeals Judge RIORDAN noted in his dissenting opinion, it makes little
difference that the communications were with the county’s elected prosecutor (as opposed
to the assistant prosecutor trying the case) because “assistant prosecutors act on behalf of
the elected county prosecutor and are supervised by him [or her] . . . .” People v Doyle,
159 Mich App 632, 644; 406 NW2d 893 (1987). For these reasons, Judge Bakker’s ex
parte communications with Prosecutor Koch violated Canon 3(A)(4) of the Michigan Code
of Judicial Conduct. And contrary to the Court of Appeals majority, I agree with the lead
opinion that the communications were not done for administrative purposes. I also agree
with the lead opinion that Judge Bakker should have known that grounds for her
disqualification existed under MCR 2.003(C)(1)(b)(ii) and Canon 2(A), and that Canon
3(C) 1 compelled the judge to raise the obligation to recuse sua sponte.
1
Canon 3(C) provides, in relevant part:
5
When defendant moved for a new trial, Judge Baillargeon granted defendant’s
motion. He found that the misconduct of Judge Bakker created an appearance of
impropriety and ruled that “the verdict has resulted in a miscarriage of justice” for purposes
of MCR 6.431(B). 2 A key issue I feel we must address is at what point an appearance of
impropriety created by judicial misconduct results in a miscarriage of justice in a criminal
case.
B. A JUDGE’S FAILURE TO RECUSE WHEN REQUIRED BY THE CANONS IS
NOT A PURELY PROCEDURAL ERROR
The primary disagreement between my view and that set forth in the lead opinion
involves how to determine whether a miscarriage of justice exists and, if it exists, whether
a new trial or resentencing can be required, even if the conduct at issue does not violate the
accused’s constitutional rights. Had the misconduct here tainted the jury’s deliberation or
demonstrated that Judge Bakker harbored actual prejudice toward defendant, then finding
a miscarriage of justice would be uncontroversial. Such issues implicate an accused’s right
to due process and a trial before an impartial jurist. While this case presents more nuanced
The judicial duties of a judge take precedence over all other activities.
Judicial duties include all the duties of office prescribed by law. In the
performance of these duties, the following standards apply:
* * *
C. Disqualification:
A judge should raise the issue of disqualification whenever the judge
has cause to believe that grounds for disqualification may exist under MCR
2.003(C).
2
MCL 769.26 also typically makes relief in a criminal case contingent on a finding that a
“miscarriage of justice” resulted from the alleged error.
6
facts, the situation is no less concerning. My focus here is to explain my disagreement
with the lead opinion’s application of MCL 769.26 and why I believe the Liljeberg test
provides the best framework for application of the harmless error standard for cases
involving nonconstitutional errors that create an appearance of impropriety.
Michigan’s statute regarding appellate review of criminal cases states as follows:
No judgment or verdict shall be set aside or reversed or a new trial be
granted by any court of this state in any criminal case, on the ground of
misdirection of the jury, or the improper admission or rejection of evidence,
or for error as to any matter of pleading or procedure, unless in the opinion
of the court, after an examination of the entire cause, it shall affirmatively
appear that the error complained of has resulted in a miscarriage of justice.
[MCL 769.26 (emphasis added).]
The statute’s plain language, in short, provides that there are very limited reasons to disturb
a verdict or sentence in a criminal case on appeal unless there has been a miscarriage of
justice. MCL 769.26 applies to three categories of potential errors in a criminal matter: (1)
misdirection of the jury, (2) the improper admission or rejection of evidence, or (3) an error
as to any matter of pleading or procedure. This matter does not concern the alleged
misdirection of the jury because the claimed error here does not involve a jury instruction.
Nor has defendant challenged on appeal the propriety of the trial court’s admission or
rejection of evidence. MCL 769.26 is thus applicable only if this appeal concerns an “error
as to any matter of pleading or procedure.” I do not believe the ex parte e-mails constituted
an error of pleading or procedure.
Significantly, the improper ex parte e-mail communications did not relate to how
the charges were pleaded, such as an error in the charging documents or selection of venue.
The communications also did not relate to the accused entering a plea of guilty or no
7
contest. Accordingly, there is no basis to conclude that this appeal concerns an error as to
a matter of the pleadings in this case. That leaves the possibility of an error of procedure
and whether the error resulted in a miscarriage of justice.
While not stated explicitly, the lead opinion implicitly concludes that the Judge
Bakker’s failure to recuse when required to do so was a mere error of procedure for
purposes of MCL 769.26. See ante at 14 (“This preserved, nonconstitutional error requires
us to consider whether defendant was entitled to relief under MCL 769.26.”). But just as I
do not think the error was one of pleading, I also do not think the error was one of
procedure. A judge’s failure to adhere to the appearance of impropriety standard in Canon
2(A) is grounds for disqualification of a judge from presiding over a case under MCR
2.003(C)(1)(b)(ii) (“Disqualification of a judge is warranted for reasons that include, but
are not limited to, the following: . . . The judge, based on objective and reasonable
perceptions, has . . . failed to adhere to the appearance of impropriety standard set forth in
Canon 2 of the Michigan Code of Judicial Conduct.”). If an error of procedure is
understood as any failure to follow the traditional or established way of doing things under
a set of rules or guidelines, then any failure of a judge to recuse could be understood as a
mere procedural defect. See The Merriam-Webster Dictionary (2005) (defining
“procedure” as “a particular way of doing something”); Black’s Law Dictionary (11th ed),
p 684 (defining “procedural error” as “[a] mistake in complying with the rules or steps in
the legal process”). But this is an overly simplistic definition of “procedure” in this context.
The Court should be hesitant to classify serious or intentional lapses in judicial ethics while
presiding over a criminal trial as mere errors of procedure.
8
In a philosophical sense, the line between procedure and substance is hazy at best.
When a judge engages in conduct on the bench that creates an appearance of impropriety
and then conceals that conduct, the judge is not merely ignoring or improperly applying
procedural norms and rules. See, e.g., TT v KL, 334 Mich App 413, 433; 965 NW2d 101(2020) (“An appearance of impropriety may arise when the conduct of a judge would create
in reasonable minds a perception that the judge’s ability to carry out judicial responsibilities
with integrity, impartiality and competence is impaired.”) (quotation marks and citation
omitted). Federal courts have long observed that ex parte proceedings “can only be
justified and allowed by compelling state interests.” In re Taylor, 567 F2d 1183, 1188 (CA
2, 1977). No compelling state interest for the ex parte communications at issue has been
articulated.
In the context of improper ex parte communications between a judge and a
prosecutor, it has been observed that “not only is it a gross breach of the appearance of
justice when the defendant’s principal adversary is given private access to the ear of the
court, it is a dangerous procedure.” Haller v Robbins, 409 F2d 857, 859 (CA 1, 1969). See
also United States v Earley, 746 F2d 412, 416 (CA 8, 1984) (“An ex parte communication
between a trial court and government counsel ‘in addition to raising questions of due
process . . . involve[s] a breach of legal and judicial ethics. Regardless of the propriety of
the court’s motives in such a case . . . the practice should be discouraged since it
undermines confidence in the impartiality of the court.’ ”), quoting 8B Moore’s Federal
Practice, ¶ 43.03[2], p 43-23 (alterations in original). And, as this Court has observed, “[a]t
the very least, participation in ex parte communications will expose the judge to one-sided
argumentation, which carries the attendant risk of an erroneous ruling on the law or facts.
9
At worst, ex parte communication is an invitation to improper influence if not outright
corruption.” Grievance Administrator v Lopatin, 462 Mich 235, 263; 612 NW2d 120(2000). By their very nature, ex parte communications raise the specter of an appearance
of impropriety. The risk is lower in degree where de minimis or nonsubstantive
communications are concerned, but the risk persists. I agree with the lead opinion that the
ex parte communications at issue were not de minimis and were substantive in nature. The
e-mails commented directly upon police-gathered evidence and policy testimony—the
very crux of the case, given the delayed disclosure and general lack of physical evidence.
Another reason to be hesitant about classifying the failure to recuse as a mere
procedural error is that all of the grounds for recusal listed under MCR 2.003(C) require
some substantive, nonprocedural basis before recusal is required. 3 Some of the grounds
3
MCR 2.003(C) states, in relevant part, as follows:
(1) Disqualification of a judge is warranted for reasons that include,
but are not limited to, the following:
(a) The judge is biased or prejudiced for or against a party or attorney.
(b) The judge, based on objective and reasonable perceptions, has
either (i) a serious risk of actual bias impacting the due process rights of a
party as enunciated in Caperton v Massey, 556 US 868; 129 S Ct 2252; 173
L Ed 2d 1208 (2009), or (ii) has failed to adhere to the appearance of
impropriety standard set forth in Canon 2 of the Michigan Code of Judicial
Conduct.
(c) The judge has personal knowledge of disputed evidentiary facts
concerning the proceeding.
(d) The judge has been consulted or employed as an attorney in the
matter in controversy.
(e) The judge was a partner of a party, attorney for a party, or a
member of a law firm representing a party within the preceding two years.
10
listed in MCR 2.003(C) hinge on constitutional concerns, such as the right to due process,
while others are rooted more firmly in the self-policing ethics of the legal profession and
common sense. The underlying need for recusal is substantive in nature, even if, when
considered in the abstract, the failure to recuse is a procedural misstep.
Whether an error is procedural or, as I believe to be the case here, another type of
error, the Court has long held that MCL 769.26 controls the review of preserved
nonconstitutional errors and permits reversal only where there has been a miscarriage of
justice. See People v Mateo, 453 Mich 203, 210-221; 551 NW2d 891 (1996). And as the
Court clarified in People v Lukity, 460 Mich 484, 495; 596 NW2d 607 (1999):
[W]e reiterate that [MCL 769.26] controls judicial review of preserved,
nonconstitutional error. [MCL 769.26] places the burden on the defendant
to demonstrate that “after an examination of the entire cause, it shall
affirmatively appear that the error asserted has resulted in a miscarriage of
justice.” We agree with the Mateo Court’s holding that reversal is only
(f) The judge knows that he or she, individually or as a fiduciary, or
the judge’s spouse, parent or child wherever residing, or any other member
of the judge’s family residing in the judge’s household, has more than a de
minimis economic interest in the subject matter in controversy that could be
substantially impacted by the proceeding.
(g) The judge or the judge’s spouse, or a person within the third degree
of relationship to either of them, or the spouse of such a person:
(i) is a party to the proceeding, or an officer, director, or trustee of a
party;
(ii) is acting as a lawyer in the proceeding;
(iii) is known by the judge to have a more than de minimis interest
that could be substantially affected by the proceeding;
(iv) is to the judge’s knowledge likely to be a material witness in the
proceeding.
11
required if such an error is prejudicial and that the appropriate inquiry
“focuses on the nature of the error and assesses its effect in light of the weight
and strength of the untainted evidence.” The object of this inquiry is to
determine if it affirmatively appears that the error asserted “undermine[s] the
reliability of the verdict.” [Citations omitted.]
Given this Court’s longstanding precedent and the text set forth in MCL 769.26,
MCL 769.26 must guide this Court’s review of Judge Bakker’s failure to recuse here. This
means that the defendant is not entitled to relief unless a miscarriage of justice has occurred.
C. A MORE NUANCED UNDERSTANDING OF “MISCARRIAGE OF JUSTICE”
This Court has historically required a defendant making a “miscarriage of justice”
claim in a criminal proceeding to set forth proof of potentially outcome determinative
prejudice that undermined the reliability of the verdict. See Lukity, 460 Mich at 492, 495-
496; Mateo, 453 Mich at 210-211. But the phrase “miscarriage of justice” appears to have
different meanings in different contexts.
When the failure to recuse results in a preserved constitutional error, then the review
of that error, despite the plain language of MCL 769.26, is not subject to a harmless error
review standard. Or it is subject to harmless error review but a miscarriage of justice is
presumed. 4 For example, this Court has explained that preserved constitutional errors that
result in a “ ‘structural defect in the constitution of the trial mechanism, which defies
4
Cf. People v Carines, 460 Mich 750, 764; 597 NW2d 130 (1999) (holding that the “plain
error rule” applies to “unpreserved claims of constitutional error”). See also, e.g., Cain v
Dep’t of Corrections, 451 Mich 470, 494-502, 508-518; 548 NW2d 210 (1996) (analyzing
arguments for and against mandatory recusal under MCR 2.003(C)(1)(a) and (b)(i)); id. at
513 n 48 (“We acknowledge there may be situations in which the appearance of
impropriety on the part of a judge or decisionmaker is so strong as to rise to the level of a
due process violation. However, this case does not present such a situation.”); Crampton
v Dep’t of State, 395 Mich 347, 351; 235 NW2d 352 (1975).
12
analysis by “harmless-error” standards,’ ” requires automatic reversal. 5 People v Anderson
(After Remand), 446 Mich 392, 404-405; 521 NW2d 538 (1994), quoting Arizona v
Fulminante, 499 US 279, 309; 111 S Ct 1246; 113 L Ed 2d 302 (1991) (brackets omitted).
Deprivation of the right to an “impartial judge” has been explicitly described as constituting
a structural constitutional error. Accordingly, reversal is required regardless of prejudice
under two of the most prominent grounds for mandatory recusal. See MCR 2.003(C)(1)(a)
(“The judge is biased or prejudiced for or against a party or attorney.”) and MCR
2.003(C)(1)(b)(i) (“The judge, based on objective and reasonable perceptions, has . . . a
serious risk of actual bias impacting the due process rights of a party as enunciated in
Caperton v Massey, 556 US 868; 129 S Ct 2252; 173 L Ed 2d 1208 (2009)[.]”). These
standards embody a constitutional requirement for recusal, and while judges are presumed
not to be biased against a party, when bias is proven or the risk of bias too great, then
recusal is mandated to ensure that the litigants receive a fair trial before a neutral arbiter.
The challenge here is that the use of a traditional harmless error review standard
does not fit well with application of most of the nonconstitutional reasons for recusal set
5
Contrary to structural constitutional errors, which require a court to “automatically
reverse,” there are “trial errors that ‘occur during the presentation of the case to the jury,
and which may therefore be quantitatively assessed in the context of other evidence
presented in order to determine whether its admission was harmless beyond a reasonable
doubt.’ ” Anderson, 446 Mich at 405-406, quoting Fulminante, 499 US at 307-308
(brackets omitted). “This requires the beneficiary of the error to prove, and the court to
determine, beyond a reasonable doubt that there is no ‘reasonable possibility that the
evidence complained of might have contributed to the conviction.’ ” Anderson, 446 Mich
at 406, quoting Chapman v California, 386 US 18, 23; 87 S Ct 824; 17 L Ed 2d 705 (1967)
(quotation marks omitted). A judge’s failure to recuse when ethically but not
constitutionally required to do so is not an error in the presentation of a case to the jury,
raising the question of how to categorize such an error.
13
forth in MCR 2.003(C)(1). For example, how could a criminal defendant or civil litigant
prove that the failure to recuse was harmful where the judge has personal knowledge of
disputed evidentiary facts, but the parties have no way to know this, and the judge has not
informed the parties of such knowledge? The need for recusal might seem obvious and
overwhelming under such circumstances, but what happens if a judge fails to recuse when
it is obviously required, and a conviction results? Must the verdict stand absent a showing
of prejudice that affected the fact-finder or a clear causal link to an adverse ruling earlier
in the proceedings? The lead opinion says yes. But I disagree.
In this regard, I find it compelling that this Court has begun to rethink its historically
rigid understanding of the “miscarriage of justice” standard set forth in MCL 769.26. And
this presumably implicates MCR 6.431(B) (motions for new trials), which also uses the
“miscarriage of justice” terminology. Last term, a majority of this Court held that the
improper denial of a criminal defendant’s peremptory challenge—even if there was no
evidence of prejudice—is an error akin to a structural constitutional error such that “leaving
defendants without a remedy in this scenario would be a miscarriage of justice for purposes
of MCL 769.26.” People v Yarbrough, 511 Mich 252, 271; 999 NW2d 372 (2023).
In Yarbrough, the judge failed to follow statutory and court rule requirements; there
was no constitutional error. But like structural constitutional errors, requiring an aggrieved
party to prove the harm caused by the erroneous denial of a peremptory challenge during
voir dire is a near-impossible feat; the nature of the error defies harmless error review.
Yarbrough observed that “there is simply no precedent to suggest that an error must be
constitutional in nature in order to be subject to automatic reversal,” noting that “Michigan
caselaw has recognized that the appropriate remedy for the erroneous denial of a
14
peremptory challenge is automatic reversal even though this right is not constitutionally
mandated.” Id. at 268. Rather, the Court held, “[t]he appropriate remedy for an error is
determined not by the source of the legal principle, but by the impact of the error and its
redressability.” Id. (emphasis added).
Similar principles are at issue here. As the United States Supreme Court has noted,
“[t]he Due Process Clause demarks only the outer boundaries of judicial disqualifications,”
Caperton, 556 US at 889 (quotation marks and citation omitted), and recusal “questions
are, in most cases, answered by common law, statute, or the professional standards of the
bench and bar,” Bracy v Bramley, 520 US 899, 904; 117 S Ct 1793; 138 L Ed 2d 97 (1997).
Nothing in MCR 2.003 relieves a judge of their obligation to raise grounds for recusal sua
sponte under Canon 3(C), and when recusal is required by MCR 2.003 but does not occur,
there should be a remedy for the violation. Where a judge’s conduct on the bench involves
improper ex parte communications with a third party about a case the judge is presiding
over, but those communications are never revealed to the jury or the parties, the potentially
affected litigants will almost never be able to prove that the error was anything but
harmless.
Under such circumstances it also would be impossible for a litigant to comply with
the timelines mandated by MCR 2.003(D) for filing a motion seeking recusal of a judge,
which are relatively short—often within 14 or 28 days of the triggering event. 6 Even when
6
MCR 2.003(D)(1)(a) provides that “[t]o avoid delaying trial and inconveniencing the
witnesses, all motions for disqualification must be filed within 14 days of the discovery of
the grounds for disqualification. If the discovery is made within 14 days of the trial date,
the motion must be made forthwith.” In this case, the defendant found out about the ex
parte communications nearly 12 months after the jury rendered its verdict.
15
a motion is filed and a request for recusal is denied, it remains a near-impossible feat to
prove whether that denial had an effect on the final judgment.
While I agree with the analysis in Part II(B) of the lead opinion explaining that the
grounds for recusal in this case do not rise to the level of violating defendant’s
constitutional rights, I do not agree that the Court’s only option for reviewing the effects
of unethical judicial conduct on a criminal trial is employing a traditional harmless error
standard. That is not to say that every failure of a judge to recuse when required should
result in upsetting a final judgment, especially where there has been a conviction by a jury
of one’s peers. I do not advocate for a rule of automatic reversal. But to force the analysis
through a traditional harmless error lens reflects, in my view, an overly myopic
understanding of the concept of “miscarriage of justice” and effectively leaves aggrieved
parties (and the public) without a remedy despite undisputed misconduct by a judge.
Instead, I would draw from the logic of our decision in Yarbrough and the decision
in Liljeberg to analyze the matter further and determine whether a miscarriage of justice
occurred under MCL 769.26 and MCR 6.431(B). And, if indeed a miscarriage of justice
occurred because of an appearance of impropriety, then we would need to determine the
proper remedy. Significantly, the United States Supreme Court has crafted a standard to
determine when relief should be granted if a judge improperly refuses to recuse under
circumstances that create an appearance of impropriety. See Liljeberg, 486 US 847. I
believe this standard can and should be incorporated into this Court’s understanding of
16
what constitutes a “miscarriage of justice” under circumstances such as those this case
presents. 7
III. THE LILJEBERG TEST IS APPROPRIATE FOR ASSESSING IMPROPER EX
PARTE COMMUNICATIONS BETWEEN A JUDGE AND A NONJUROR
THIRD PARTY
In Liljeberg, the United States Supreme Court was faced with determining whether
a late-discovered violation of the federal recusal statute, 28 USC 455(a), 8 required vacating
a final judgment and ordering new proceedings. Liljeberg, 486 US at 850. This situation
7
Nothing in my opinion should be read as disparaging the multitiered test applicable for
ex parte communications between a trial court and a jury that was articulated in People v
France, 436 Mich 138; 461 NW2d 621 (1990). Nor does the failure to recuse after a
judge’s conduct or actions have created an appearance of impropriety rise to the same level
as when a judge improperly injects themselves into a criminal trial such that the veil of
impartiality is destroyed, creating a structural constitutional error. See People v Stevens,
498 Mich 162; 869 NW2d 233 (2015).
8
At the time Liljeberg was decided, 28 USC 455 provided in relevant part:
(a) Any justice, judge, or magistrate of the United States shall
disqualify himself in any proceeding in which his impartiality might
reasonably be questioned.
(b) He shall also disqualify himself in the following circumstances:
* * *
(4) He knows that he, individually or as a fiduciary, or his spouse or
minor child residing in his household, has a financial interest in the subject
matter in controversy or in a party to the proceeding, or any other interest
that could be substantially affected by the outcome of the proceeding.
* * *
(c) A judge should inform himself about his personal and fiduciary
financial interests, and make a reasonable effort to inform himself about the
personal financial interests of his spouse and minor children residing in his
household.
17
arose from a lawsuit commenced in 1981 to determine the ownership of a corporation (St.
Jude Hospital of Kenner, Louisiana (St. Jude)), which the plaintiff had formed to obtain
regulatory approvals needed to construct and operate a hospital. Id. at 850, 852. Over the
course of several years, the plaintiff had been negotiating with at least two parties regarding
competing locations and plans for the hospital and related aspects of the project. Id. at 853.
One of those parties was Loyola University, which owned a large tract of land that could
serve as a hospital site and provide other benefits to the university; the other was a
predecessor to the defendant company, which was interested in a different formulation of
the project. Id.
The outcome of the litigation concerning St. Jude was material to Loyola
University’s financial interest in the project. Significantly, while the lawsuit was pending
before the district court judge, Judge Robert Collins, the plaintiff and Loyola University
executed an agreement to sell the real estate to the plaintiff for more than $6,000,000. Id.
at 856-857. This was problematic because, during the same period, Judge Collins was a
member of the board of trustees for Loyola University and had participated in and voted
during meetings where the university’s interest in the hospital project was at issue. Id. at
856-858.
While Judge Collins knew of the university’s interest in St. Jude in 1980 and 1981
from having attended board meetings during which negotiations regarding the hospital
project were discussed days before the Liljeberg case was filed, he claimed to have
forgotten about this before the bench trial in January 1982. Id. at 856-857. Judge Collins
did not attend the meeting at which Loyola University agreed to sell real property to the
defendant, but the meeting minutes arrived by mail a day or two before the judge entered
18
the final judgment in the case, and that mail sat unopened. Id. at 855, 858. But within a
few days of filing the final judgment on March 16, 1982, Judge Collins indisputably
regained actual knowledge of the university’s interest in St. Jude and the pending litigation
when he read the meeting minutes in preparation for the next board meeting. Id. at 851,
857-858. Despite this, Judge Collins refused to sua sponte recuse himself under 28 USC
455(a). When a party later learned of the conflict and moved to vacate the judgment under
FR Civ P 60(b)(6), Judge Collins denied the motion. Id. at 850-851. On appeal, the United
States Court of Appeals for the Fifth Circuit held that Judge Collins should have recused,
and it vacated the judgment. Id. at 858. The matter was then appealed to the United States
Supreme Court.
In deciding whether violation of the recusal statute warranted vacatur of the final
judgment, the Liljeberg Court noted that “[t]he judge’s lack of knowledge of a
disqualifying circumstance may bear on the question of remedy, but it does not eliminate
the risk that ‘his impartiality might reasonably be questioned’ by other persons.” Id. at
859, quoting 28 USC 455(a). And the Court made clear that while a judge cannot be
required to do the impossible—recuse based on facts they do not know—a judge can be
required to recuse and possibly take other action, such as vacating a judgment, to remedy
the potential for harm once the conflict is learned of. Liljeberg, 486 US at 861. After
concluding that the recusal statute had been violated, the Court stated:
As in other areas of the law, there is surely room for harmless error
committed by busy judges who inadvertently overlook a disqualifying
circumstance. There need not be a draconian remedy for every violation of
§ 455(a). It would be equally wrong, however, to adopt an absolute
prohibition against any relief in cases involving forgetful judges. [Id. at 862.]
19
The Liljeberg Court held that while 28 USC 455 was silent as to the remedy for a
judge’s failure to recuse, FR Civ P 60(b)(6) authorized relief from a final judgment “upon
such terms as are just” if a motion is filed within a reasonable time and is not premised on
one of the enumerated grounds. Id. at 863. The Court stated that the rule “provides courts
with authority ‘adequate to enable them to vacate judgments whenever such action is
appropriate to accomplish justice,’ while also cautioning that it should only be applied in
‘extraordinary circumstances.’ ” Id. at 864 (citations omitted). In determining whether a
final judgment should be vacated for violation of the recusal statute, the Court set forth
three factors to be weighed: “[1] the risk of injustice to the parties in the particular case,
[2] the risk that the denial of relief will produce injustice in other cases, and [3] the risk of
undermining the public’s confidence in the judicial process.” Id. Moreover, courts “must
continuously bear in mind that ‘to perform its high function in the best way[,] justice must
satisfy the appearance of justice.’ ” Id., quoting In re Murchison, 349 US 133, 136; 75 S
Ct 623; 99 L Ed 942 (1955).
It is noteworthy that in affirming the Fifth Circuit’s holding that vacatur was the
proper remedy, the Liljeberg majority emphasized four facts “that might reasonably cause
an objective observer to question Judge Collins’ impartiality.” Liljeberg, 486 US at 865.
First, the Court found it “remarkable” that a judge who had been on the university’s board
for three to four years at the time the litigation started would have completely forgotten
about the university’s interest in having a hospital constructed on the real property at issue,
especially considering the value of the project and the number of formal actions the board
would have had to approve. Id. Second, the Court found it concerning that while it
appeared that the board was monitoring the progress of the litigation and Judge Collins
20
regularly attended meetings, no one on the board called the judge’s attention to the obvious
conflict of interest. Id. at 866. Third, it was “remarkable—and quite inexcusable—that
Judge Collins failed to recuse himself” from postjudgment proceedings once he had actual
knowledge of the conflict. Id. A “full disclosure at that time would have completely
removed any basis for questioning the judge’s impartiality and would have made it possible
for a different judge to decide whether the interests—and appearance—of justice would
have been served by a retrial.” Id. A second two-day bench trial or evidentiary hearing
would have been less burdensome than what resulted from the failure to recuse. Id. Fourth,
when the motion to vacate was filed, Judge Collins “did not acknowledge that he had
known about the University’s interest both shortly before and shortly after the trial.” Id.
at 867.
The Liljeberg Court went on to note that the enforcement of § 455 under such
circumstances “may prevent a substantive injustice in some future case by encouraging a
judge or litigant to more carefully examine possible grounds for disqualification and to
promptly disclose them when discovered.” Id. at 868. Turning to the fairness to the
litigants, the Court concluded that there was a greater risk of unfairness in upholding the
judgment than in having a new judge take a fresh look, and neither party had shown a
special hardship in reliance on the original judgment. Id.
In the decades since Liljeberg was decided, application of its three-factor test has
not been limited to civil matters. Quite recently, the Liljeberg test was applied by the
United States Court of Appeals for the Seventh Circuit after a federal district court judge
engaged in misconduct very similar to that which occurred in this case.
21
In United States v Atwood, 941 F3d 883 (CA 7, 2019), the defendant pleaded guilty
to three counts of federal drug crimes and was sentenced by federal district court judge
Colin Bruce. It later came to light that while the defendant’s case was pending, Judge
Bruce had “engaged in extensive ex parte communication with the prosecuting U.S.
Attorney’s Office about other cases.” 9 Id. at 884. Judge Bruce had previously worked as
a prosecutor in the same office for many years. Id. The communications varied in content.
Some were substantive and concerned cases pending before Judge Bruce, such as an e-mail
in which Judge Bruce “expressed exasperation that the novice prosecutor’s weak cross-
examination had turned the case ‘from a slam-dunk for the prosecution to about a 60-40
for the defendant . . . .’ ” Id. Other communications were more ministerial in nature, but
they also showed Judge Bruce “cheering on Office employees and addressing them by
nicknames.” Id. “At other times, Judge Bruce wrote to prosecutors in the Office to
congratulate and thank them for persuading [the Seventh Circuit] to affirm his decisions.
Judge Bruce never explicitly mentioned Atwood’s case in an ex parte email.” Id.
The government in Atwood conceded that the communications “gave the appearance
that [Judge Bruce] was biased in the government’s favor but argue[d] that the error was
harmless.” Id. at 885. The appellate court disagreed and held that pursuant to Liljeberg,
9
While Atwood was pending on appeal, the Seventh Circuit Judicial Council issued an order
admonishing Judge Bruce and ordering him to complete additional ethics training and
“remain unassigned to any matters involving the U.S. Attorney’s Office in the Central
District of Illinois until September 1, 2019.” In re Complaints Against Dist Judge Colin S
Bruce, order of the Judicial Council of the Seventh Circuit, entered May 14, 2019 (Complaint
Nos. 07-18-90053 and 07-18-90067), available at <http://www.ca7.uscourts.gov/judicial-
conduct/judicial-conduct_2018/07_18-90053_and_07-18-90067.pdf> (accessed April 25,
2024) [https://perma.cc/79SG-XYGT].
22
the defendant was entitled to be resentenced before a different judge. Id. at 885-886. The
court explained that there would be a “real risk of unfairness” to the defendant if his
sentence were allowed to stand because even though the sentence was calculated using the
federal sentencing guidelines, judges exercise a great deal of discretion in sentencing that
could be or appear to be influenced by personal biases. Id. at 885. Conversely, a remand
for resentencing presented little risk of unfairness to the government because while there
were some administrative costs, those costs would not impose a special hardship. Id. The
panel further stated that “[a]s in Liljeberg, we think that enforcing [28 USC] 455(a) in this
case ‘may prevent a substantive injustice in some future case’—here, by encouraging
judges to exercise caution in their communications.” Id., quoting Liljeberg, 486 US at 868.
Concerning the risk of harm to public confidence:
In sentencing, the most significant restriction on a judge’s ample discretion
is the judge’s own sense of equity and good judgment. When those qualities
appear to be compromised, the public has little reason to trust the integrity of
the resulting sentence. The government has conceded that Judge Bruce
compromised his appearance of impartiality. Allowing Atwood’s sentence
to stand would undermine the public’s confidence in the fairness of this
sentence and in the impartiality of the judiciary. [Atwood, 941 F3d at 886.]
The Liljeberg test was also applied in other cases involving Judge Bruce. In United
States v Williams, 949 F3d 1056 (CA 7, 2020), Judge Bruce presided over the defendant’s
trial but, unlike in Atwood, the defendant was sentenced by another judge. A significant
distinction from Atwood for purposes of the first Liljeberg factor (risk of injustice to the
parties) was that the defendant in Williams had not shown that Judge Bruce made any
significant decisions that involved broad discretion, and “judges generally have more
discretion over sentencing than the outcome of a jury trial.” Id. at 1064. Additionally, in
23
Williams, all pretrial and trial rulings “were not controversial or contested and generally
pertained to routine matters,” and none were challenged on appeal. Id. The panel also
noted that since Atwood had been decided, Judge Bruce had been investigated and
reprimanded for his misconduct, which weighed against granting relief under the second
Liljeberg factor (risk that denial of relief will result in injustice in other cases). Id. at 1065.
The third Liljeberg factor (risk of undermining the public’s confidence) was a close call,
but because Judge Bruce did not make significant discretionary rulings that could have
influenced the outcome of the jury conviction, no actual bias was demonstrated. Id. at
1065-1066. Further, the panel concluded that overturning a jury verdict could have a
negative impact on the public’s confidence in the judiciary and, therefore, this factor did
not favor granting relief. Id. Accordingly, the court held that Judge Bruce’s failure to
recuse himself, despite being required to do so under 28 USC 455(a), was a harmless error.
Id. at 1066.
In United States v Orr, 969 F3d 732 (CA 7, 2020), another defendant whose trial
was presided over by Judge Bruce argued that he was entitled to a new trial because of the
judge’s ex parte communications in violation of 28 USC 455(a). Unlike the defendant in
Williams, this defendant specifically “challenge[d] three seemingly discretionary decisions
by Judge Bruce: the denial of the motion to suppress, the admission of drug evidence under
Federal Rule of Evidence 404(b), and the allowance of cross-examination questions about
Orr’s prior felony conviction.” Id. at 739. While the panel determined that no reasonable
jurist could have reached a different result as to the motion to suppress, the other two
rulings were closer discretionary calls and aided the prosecution by permitting additional
circumstantial evidence into the trial. Id. at 740-741.
24
The first Liljeberg factor favored the defendant. While the government would incur
expenses in retrying the case, the trial was short, the theory of the case was straightforward,
and the risk of injustice to the defendant that the personal biases of the trial judge influenced
the outcome was higher. Id. at 741. As to the second factor, the court noted that
considering Atwood and the fact that Judge Bruce had been investigated and reprimanded
already, vacating the defendant’s conviction would provide little additional benefit toward
preventing injustice in other cases. Id. However, the panel found that the third factor still
favored granting a new trial because, even though the defendant was convicted by a jury
of his peers, the discretionary decisions made by Judge Bruce were consequential and
bolstered the prosecution’s case, which already rested primarily on circumstantial
evidence. Id. at 742. “Given these discretionary rulings, upholding [defendant] Orr’s
conviction may damage the public’s confidence in the impartiality of the judiciary.” 10 Id.
Considering the reasoned application of Liljeberg in the criminal cases previously
discussed, I disagree with the lead opinion’s statement that the Liljeberg test is “vague to
the point of futility.” 11 Ante at 17. While the test takes a different approach toward a
10
At least two other criminal matters over which Judge Bruce presided resulted in similar
outcomes. See, e.g., Shannon v United States, 39 F4th 868, 883-888 (CA 7, 2022) (granting
the defendant resentencing before a different judge because of Judge Bruce’s ex parte
communications and concerning commentary, but declining to grant a new trial because
like in Williams no evidence of bias existed and the judge did not make significant,
controversial discretionary rulings during the trial); United States v Gmoser, 30 F4th 646,
649 (CA 7, 2022) (declining to grant relief because “Gmoser’s trial was not even arguably
affected by any improper communication or any material discretionary decision. As
Gmoser’s sentence was imposed by a different judge, we do not see a good reason to start
from scratch”).
11
Contrary to the conclusions reached by Justice BOLDEN, I believe the series of decisions
in which the Seventh Circuit applied Liljeberg to similar judicial misbehavior demonstrates
that the test works quite well when applied to criminal matters in which a judge’s conduct
25
narrow class of nonconstitutional errors than this Court historically has when determining
whether an error has resulted in a “miscarriage of justice” under MCR 769.26 or MCR
6.431(B), it is compatible with the values these rules seek to preserve. The previously
discussed Seventh Circuit decisions set forth a workable framework that protects the
integrity of the judiciary and the public. There is no need to attempt to confine the concept
of “miscarriage of justice” in a one-size-fits-all manner. Rather than focusing purely on
whether an error was so significant as to undermine the reliability of the verdict in a
criminal case, the Liljeberg test allows for consideration of fairness to the parties and
whether the public can be confident in the overall fairness of the criminal proceeding.
There is a “compelling interest in preserving the integrity of the judiciary. . . . The
state’s interest in the integrity of the judiciary extends to preserving public confidence in
the judiciary. The appearance of fairness and impartiality is necessary to foster the people’s
willingness to accept and follow court orders.” In re Chumra, 461 Mich 517, 534-535; 608
NW2d 31 (2000). See also Landmark Communications, Inc v Virginia, 435 US 829, 848;
98 S Ct 1535; 56 L Ed 2d 1 (1978) (Stewart, J., concurring) (“There could hardly be a
higher governmental interest than a State’s interest in the quality of its judiciary.”). The
creates an appearance of impropriety that does not rise to the level of a constitutional
violation. The flexible Liljeberg test is not a replacement for the more rigid standards that
apply where a judge’s conduct leads to the violation of a criminal defendant’s constitutional
rights. While Liljeberg is nonbinding in the sense that it concerns a failure to recuse under
28 USC 455, I find the decision and the appellate decisions applying it persuasive
considering the similarities between Michigan’s recusal rules in MCR 2.003(C) and the
analogous judicial canons concerning ex parte communications and avoiding an
appearance of impropriety. See Abela v Gen Motors Corp, 469 Mich 603, 607; 677 NW2d
325 (2004). The Seventh Circuit decisions discussed are highly persuasive, too, given their
application of Liljeberg to facts that are nearly identical to those before us in this case.
26
law must recognize that when judicial misconduct seriously undermines public confidence
in a criminal proceeding but does not amount to a constitutional error, some degree of relief
might be warranted even if it cannot be determined that the judge’s conduct had a definitive
effect on the outcome of the proceedings. Moreover, as Liljeberg recognized, not every
instance of judicial misconduct creating an appearance of impropriety will be significant
enough to justify setting aside an otherwise valid judgment. A case-by-case inquiry is
necessary to determine whether a particular set of facts warrants granting relief, and the
Liljeberg test can be a useful tool in determining whether a miscarriage of justice has
occurred.
The previously discussed decisions in Atwood, Williams, and Orr demonstrate that
Liljeberg can be applied in a rational and measured way to a narrow class of
nonconstitutional errors in criminal proceedings. Where there has been a jury conviction
and there is no evidence that the jury was aware of the presiding judge’s ethical lapses, it
is useful to consider whether the offending judge exercised judicial discretion that might
have affected pretrial and trial rulings or the sentence that was imposed following a
conviction. In a criminal jury trial, the primary areas in which a presiding judge exercises
discretion are its rulings on pretrial and trial motions and during sentencing (assuming a
mandatory minimum sentence is not applicable). This focus is appropriate because the
most obvious harm caused by improper ex parte communications between a judge and a
prosecutor is that the public might reasonably question whether the judge is biased in favor
of the prosecution, even if there is no evidence of actual bias (which, of course, would
result in a structural constitutional problem).
27
IV. THE CURRENT RECORD REQUIRES A REMAND TO APPLY THE
LILJEBERG FRAMEWORK
Having concluded that the ex parte communications at issue in this case created an
appearance of impropriety and that the Liljeberg test should guide the outcome of this
appeal, the next question is whether defendant is entitled to relief. If the Court adopted my
view—an augmentation of Michigan’s harmless error standard for appearances of
impropriety that are nonconstitutional errors—then the next step would be to remand the
case to a judge other than Judge Bakker to apply the Liljeberg framework.
Defendant emphasizes that he was facing the possibility of a 25-year mandatory
minimum sentence for CSC-I, that Judge Bakker’s concealment of the ex parte
communications deprived him of the opportunity to respond or to move for a mistrial, and
that the ex parte communications highlighted weaknesses in the prosecution’s case and
could have altered the prosecuting attorney’s trial strategy. Defendant, therefore, argues
that the error here is akin to a structural error and that he is entitled to a new trial.
The prosecution emphasizes that even if Judge Bakker’s communications were
improper, the jury never learned of them, the time stamps on the e-mails and the trial
transcript demonstrate that the presentation of evidence and arguments was not altered
because of the e-mails, there is no evidence that the assistant prosecuting attorney trying
the case was even aware of the e-mails, and the ex parte communications do not
demonstrate actual bias. The prosecution, therefore, argues that no structural error
occurred, Judge Bakker’s ex parte communications were harmless, and defendant is not
entitled to relief.
28
Turning to the Liljeberg factors, the first consideration is weighing the risk of
injustice to the parties in the case that would be caused by granting or denying relief. Given
that defendant was convicted by a jury of his peers and the jury was unaware of the
improper ex parte communications, it would be relevant to consider the degree to which
Judge Bakker exercised her discretion during these criminal proceedings and whether those
rulings were routine in nature or required more judicial discretion. If more discretionary
in nature, then the trial court would need to examine who benefited from those rulings.
While both parties touch on this briefly, the parties deserve an opportunity to examine this
more closely on remand. Further, on remand, a trial court would need to consider the extent
to which ordering a new trial would hinder the ability of the prosecution to present the case
again, the costs of such a trial, and the impact upon the complainant. Other relevant
considerations are that significant time has passed, memories have likely faded, and the
availability of witnesses could be uncertain. The trial court judge would need to determine
whether this case more resembles Williams, 949 F3d at 1066 (in which the judge’s rulings
lacked discretion and the defendant was sentenced by a different judge, rendering any error
harmless), or Orr, 969 F3d at 741 (in which the Liljeberg factors supported a new trial
given that the judge’s rulings were discretionary and controversial). It could also be that
instead of a new trial, resentencing by a different judge, as was ordered in Atwood and
Shannon, is what is required to remedy the appearance of impropriety. A trial court on
remand would be in a better position to make that determination than this Court.
The second consideration under Liljeberg is the risk that denial of relief will produce
injustice in other cases. The trial court on remand may want to consider the secret nature
of the communications in this matter and how the ex parte communications came to light
29
via an unrelated third party filing a Freedom of Information Act request in 2020. Moreover,
as the amicus brief filed by Watson Township suggests, the ex parte e-mails at issue in this
case were not necessarily the only ex parte communications between Judge Bakker and
Prosecutor Koch. Media reports also indicate that an ethics complaint has been filed
against both Prosecutor Koch and Judge Bakker, 12 but given the confidential nature of such
complaints, until the relevant authority decides to initiate public disciplinary action, it is
unclear whether these complaints have been investigated or whether disciplinary action
will be pursued.
The third Liljeberg factor considers the risk of undermining the public’s confidence
in the judicial process if relief is not granted. “At the very least, participation in ex parte
communications will expose the judge to one-sided argumentation, which carries the
attendant risk of an erroneous ruling on the law or facts. At worst, ex parte communication
is an invitation to improper influence if not outright corruption.” Lopatin, 462 Mich at
263. The trial court judge on remand would need to consider whether the facts presented
pose a great risk of undermining the public’s confidence in the judicial process and the
impartiality of the judiciary. A trial court judge would also have to balance the lack of
evidence that the ex parte communication affected the jury verdict, the potential harm to
the complainant, and the risk of eroding public confidence if defendant’s convictions of
multiple counts of criminal sexual conduct were vacated solely because of an appearance
12
See Gamble, Complaints Filed Against Allegan County Judge, Prosecutor by
Challenger, Holland Sentinel (July 31, 2020), available at
<https://www.hollandsentinel.com/story/news/politics/elections/2020/07/31/complaints-
filed-against-allegan-county-judge-prosecutor-by-challenger/114347090/> (accessed
April 26, 2024) [https://perma.cc/4CQL-NNFM].
30
of impropriety. See Williams, 949 F3d at 1066 (noting that “overturning a jury verdict
based purely on the appearance of bias creates a risk that the public will lose confidence in
the judicial process”).
The balance of the Liljeberg factors could produce a different outcome than what
either defendant or the prosecution suggest. I would remand this case to a new trial court
judge for application of the Liljeberg framework.
V. CONCLUSION
For the reasons previously explained, I join the lead opinion as to Part I and Parts
II(A)(1) and (A)(2), as well as Part II(B). However, I cannot joint Part II(A)(3) or the full
judgment of the lead opinion because I believe the Liljeberg framework is consistent with
harmless error review under MCL 769.26 where the alleged error is inappropriate ex parte
communications by a trial judge that create an appearance of impropriety but do not rise to
the level of a constitutional violation. I would remand this case to a new trial court judge
to examine the harmless error standard through the Liljeberg framework. Because the lead
opinion takes a different approach and reaches a different result, I respectfully dissent.
Elizabeth M. Welch
Megan K. Cavanagh
31
STATE OF MICHIGAN
SUPREME COURT
PEOPLE OF THE STATE OF MICHIGAN,
Plaintiff-Appellee,
v No. 164133
DANIEL ALBERT LOEW,
Defendant-Appellant.
BERNSTEIN, J. (dissenting).
I respectfully dissent from the lead opinion because I do not believe that its approach
to this case properly captures the constitutional question presented here.
To start, I highlight my points of agreement with my colleagues. I agree with the
lead opinion and Justice BOLDEN that this Court need not adopt the test that the United
States Supreme Court articulated in Liljeberg v Health Servs Acquisition Corp, 486 US
847; 108 S Ct 2194; 100 L Ed 2d 855 (1988). In particular, I join the part of Justice
BOLDEN’S opinion regarding why Liljeberg is an unsuitable test for this Court to adopt.
Like Justice BOLDEN, I do not believe that we need to rely on the Michigan Code of Judicial
Conduct to resolve this case. However, unlike my colleagues, I would resolve this case on
the constitutional question alone, adopting the test that the United States Court of Appeals
for the Sixth Circuit set forth in United States v Barnwell, 477 F3d 844 (CA 6, 2007).
Because the parties have not had an opportunity to brief arguments under Barnwell, I would
remand this case to the Court of Appeals with instructions to apply the test and determine
whether defendant is entitled to a new trial. For these reasons, I respectfully dissent.
The lead opinion here considers this case under two different legal theories. The
lead opinion first explains that the trial judge should have sua sponte recused herself from
the case and that her failure to do so violated MCR 2.003(c)(1)(b)(ii). Because this theory
is premised on a court rule and not a constitutional principle, the lead opinion analyzes this
issue as a nonconstitutional error. The lead opinion recognizes that, typically, a defendant
is expected to preserve an error through objection. Defendant’s lack of objection would
thus render this an unpreserved issue. But because defendant here did not know that the
ex parte communications occurred, he could not have possibly preserved an objection at
his trial. Accordingly, the lead opinion treats the issue as preserved on the basis of the
“exceptional circumstance” presented by this case, but concludes that defendant cannot
demonstrate a miscarriage of justice as required by MCL 769.26. The lead opinion then
goes on to explain that the ex parte communications that occurred between the trial judge
and the county prosecutor raise due-process concerns. The lead opinion concludes, though,
that defendant was not deprived of due process because defendant could not establish that
the ex parte communications affected his case or that he was (1) deprived of a right to trial
by an impartial decisionmaker; or (2) deprived of a right to be present and his right to
counsel. I find the lead opinion’s analytical framework unsuitable for resolving the
question posed in this case, and would instead address defendant’s argument solely in due-
process terms.
I. THE LEAD OPINION’S NONCONSTITUTIONAL ANALYSIS
In considering whether defendant was entitled to a new trial on the basis of the trial
judge’s failure to recuse herself, the lead opinion relies on MCR 6.431(B), which governs
2
a defendant’s motion for a new trial. The rule provides that a trial court “may order a new
trial on any ground that would support appellate reversal of the conviction or because it
believes that the verdict has resulted in a miscarriage of justice.” MCR 6.431(B). The lead
opinion characterizes the trial judge’s failure to recuse herself as a nonconstitutional error,
subject to the requirement that defendant show a miscarriage of justice under MCL 769.26.
Under our caselaw, a miscarriage of justice has occurred if a reviewing court determines,
after evaluating “the effect of the error . . . by assessing it in the context of the untainted
evidence,” that “it is more probable than not that a different outcome would have resulted
without the error.” People v Lukity, 460 Mich 484, 495; 596 NW2d 607 (1999). The lead
opinion concludes that defendant could not satisfy that burden.
But defendant has maintained throughout the appellate process that the ex parte
communications between the trial judge and the county prosecutor constituted a violation
of his due-process rights. Defendant did not raise an isolated argument that the trial judge
should have recused herself. 1 Instead, defendant has consistently argued that his trial was
fundamentally unfair as a result of the ex parte communications. 2 Nonetheless, the lead
1
This makes sense. When the trial judge engaged in ex parte communications, which
would have been the basis for any plausible recusal motion, it was during defendant’s trial.
Defendant did not learn about the communications until after the trial concluded and he
had already been found guilty. Thus, it was impossible for defendant to have brought a
motion for recusal.
2
On appeal, defendant largely argued that the ex parte communications resulted in the
deprivation of his due-process right to a fair trial. The Court of Appeals majority and
dissent discussed nonconstitutional theories of the case. Our Court asked for briefing on
these issues when we granted oral argument. While I do not fault the lead opinion for
answering these questions, I find that the answers to these questions are unnecessary to
resolve the case.
3
opinion explains that because the trial judge should have known that grounds for her
disqualification might have existed under MCR 2.003(C)(1)(b)(ii) and Canon 3(C), she
should have raised the issue of her disqualification sua sponte and recused herself. Because
the trial judge did not sua sponte recuse herself, the lead opinion first considers this case
through a nonconstitutional lens, imposing the burden on defendant to demonstrate a
miscarriage of justice.
But imposing a miscarriage-of-justice standard on defendants in cases in which a
judge is alleged to have engaged in misconduct is both unfair to defendants and a
substantively futile exercise. As the lead opinion notes, a miscarriage of justice occurs
only when, but for a nonconstitutional error, it is more likely than not that the jury would
have reached a different conclusion. While there are some circumstances in which
prejudice would be presumed in this inquiry, whether and how ex parte communications
have affected a fact-finder would be nearly impossible for a defendant to prove. That is to
say, even on the rare occasion in which a judge explicitly directs a prosecutor on how to
proceed with its case-in-chief, a defendant would face an uphill battle in demonstrating that
(1) the prosecutor changed its arguments or presentation of evidence in response to those
communications; and (2) the jury’s vote was swayed by the prosecution’s new presentation.
In other words, the lead opinion’s position presents a uniquely difficult way of proving
prejudice because the ex parte communications and their effect on the prosecution are all,
by definition, outside the trial court record. I believe that by imposing this difficult burden
on defendants, the lead opinion’s resolution of this issue necessarily conflicts with the
general understanding that ex parte communications are disfavored in law. See Ayestas v
Davis, 584 US 28, 40; 138 S Ct 1080; 200 L Ed 2d 376 (2018); Hereford v Warren, 536
4
F3d 523, 530 (CA 6, 2008); United States v Carmichael, 232 F3d 510, 518 (CA 6, 2000)
(“As a general rule of thumb, in all but the most exceptional circumstances, ex parte
communications with the court are an extraordinarily bad idea. This court has not
concealed its strong disapproval of ex parte approaches in criminal cases, reasoning that
giving the government private access to the ear of the court is not only ‘a gross breach of
the appearance of justice,’ but also a ‘dangerous procedure.’ ”) (citation omitted). For these
reasons, I find it unworkable to consider a case like this though a nonconstitutional lens. 3
II. DUE PROCESS AND EX PARTE COMMUNICATIONS
I would instead address defendant’s issue under a due-process theory. Defendant’s
requests for a new trial have been by way of the constitutional arguments that underlie his
appellate claims. Thus, I would consider whether the ex parte communications between
the trial judge and the county prosecutor deprived defendant of a fair trial, such that
defendant is entitled to a new trial under a constitutional due-process inquiry.
A defendant has a constitutional right to a fair trial. People v Horton, 341 Mich
App 397, 401; 989 NW2d 885 (2022), citing US Const, Am XIV; Const 1963, art 1, § 17
(“Implicit in this [due-process] guarantee is that each criminal defendant enjoys the right
to a fair trial, and essential to a fair trial is the defendant’s right to be presumed innocent.”).
3
Because of the lack of evidence available to defendant and the high evidentiary burden
imposed on defendants who proceed with nonconstitutional claims, it would not have been
strategic for defendant to proceed on appeal with an unpreserved nonconstitutional issue.
As I will explain more fully, the Sixth Circuit has crafted a specific constitutional test that
governs consideration of cases in which a judge and a prosecutor engage in ex parte
discussions. Barnwell, 477 F3d at 851; United States v Minsky, 963 F2d 870, 874 (CA 6,
1992). Those rulings, which I would adopt, more accurately reflect the historical
understanding that ex parte communications are disfavored. Ayestas, 584 US at 40.
5
Due process affords a criminal defendant “an impartial and disinterested tribunal . . . .”
Marshall v Jerrico, Inc, 446 US 238, 242; 100 S Ct 1610; 64 L Ed 2d 182 (1980). Also
inherent in the right to a fair trial is the right to be present at all stages of the trial during
which a defendant’s absence might frustrate the fairness of the proceedings. Illinois v
Allen, 397 US 337, 338; 90 S Ct 1057; 25 L Ed 2d 353 (1970); People v Buie (On Remand),
298 Mich App 50, 56-57; 825 NW2d 361 (2012).
Here, I consider whether ex parte communications between a judge and a prosecutor
deprive a defendant of the due-process right to a fair trial, such that defendant here is
entitled to a new trial. More specifically, the trial judge in this case engaged in ex parte
communications with the county prosecutor, indicating what she believed to be
deficiencies in the prosecution’s case-in-chief. In closing arguments, the trial prosecutor
addressed these deficiencies in more detail than she did in opening arguments. 4 The jury
then returned a guilty verdict against defendant.
This Court has never before adopted a specific framework to consider the
constitutional implications of ex parte communications between a judge and a prosecutor.
Here, the lead opinion considers (1) whether defendant was denied due process;
(2) whether defendant was deprived of his right to a trial presided over by an impartial
judge and jury; and (3) whether the trial judge’s ex parte communications deprived him of
his constitutional right to be present and his right to counsel. The lead opinion concludes
that defendant has not established that he was deprived of any right largely because (1) ex
4
Perhaps the trial prosecutor would have done this regardless of whether the ex parte
communications occurred, but no one can answer that question with any degree of
certainty. Rather than ask future courts to engage in guesswork when this type of issue
arises, I would adopt a test that does away with this sort of speculation.
6
parte communications are not per se unconstitutional; (2) defendant failed to show that the
trial judge was actually biased or that there was an unconstitutionally high probability she
was actually biased; and (3) defendant failed to demonstrate that the trial prosecutor altered
her strategy in response to the trial judge’s ex parte communications with the county
prosecutor. I agree with the lead opinion that ex parte communications are not per se
unconstitutional, but my inquiry would not stop there. I find it instructive to look at how
courts have traditionally considered ex parte communications.
Justice BOLDEN finds People v France, 436 Mich 138; 461 NW2d 621 (1990),
instructive. I agree. There, this Court recognized that ex parte communications squarely
implicate the due-process right to a fair trial. See id. at 166. Instead of centering its holding
on the right to an impartial decisionmaker or the right to be present, the France Court more
generally considered the nature of ex parte communications and set out several factors to
assess that nature. The Court then established a presumption that an ex parte
communication is prejudicial when it is substantive. Id. at 143 (“A substantive
communication carries a presumption of prejudice in favor of the aggrieved party
regardless of whether an objection is raised.”). France, however, considered ex parte
communications between a judge and jury, and much of the Court’s discussion in that case
centered on the specific problems posed by such communications. I find the Sixth Circuit
decision in Barnwell, which considered ex parte communications between a judge and a
prosecutor, to be more persuasive in this context and to largely comport with this Court’s
general due-process consideration of ex parte communications in France. I would adopt
Barnwell here.
7
The lead opinion concludes that the facts in this case are “poles apart from
Barnwell.” While I recognize the factual differences between the case here and Barnwell,
I believe that the overall legal principle set forth in Barnwell is easily applied to this case
and others like it.
In Barnwell, the defendant was tried before a jury for crimes related to
misappropriating the assets of a labor union. 5 Barnwell, 477 F3d at 845. Both the
prosecution and the defendant presented their case and rested. Id. at 847. The jury began
deliberations. Id. After the jury began to deliberate, the Federal Bureau of Investigation
intercepted a phone call that discussed “the trial of ‘our friend’ downtown, and indicated
that a woman had advised them that the vote was 8 to 4 for conviction at one point, but had
shifted towards acquittal, 10 to 2.” Id. The intercepted phone call was brought to the
attention of the office of the Assistant United States Attorney (AUSA).
Having believed that the intercepted phone call was related to the Barnwell case, an
attorney in the AUSA office left a voicemail for the presiding judge, stating that he had an
important matter he had to discuss. Id. While the jury continued deliberations, the
presiding judge engaged in ex parte communications with the government a total of five
times regarding the intercepted phone call and whether the case was compromised. Id. at
850. The presiding judge also met in camera with the jury foreperson, who said “there was
never an 8 to 4 vote and the vote was 11 to 1, not 10 to 2.” Id. at 848. The presiding judge
later informed the government that “the way I’m thinking it is (sic), we are going to get a
5
Defendant Barnwell was indicted along with four codefendants. The use of “the
defendant” when discussing this case refers only to defendant Barnwell, as he was the only
defendant to raise an appeal.
8
note something like they’re hung which is the way to do it. I think it’s important not to
disclose or have anybody think about the wiretap . . . .” Id. (quotation marks and citation
omitted). It was the presiding judge’s theory that the dissenting juror must have been the
partial juror. Id. The presiding judge suggested dismissing the dissenting juror. Id. The
government wanted to take the risk and see if it could “get a conviction on some of the
defendants.” Id. (quotation marks omitted). Peculiarly, the government and the presiding
judge did not reach a consensus on a course of action. Id. Though this was the
government’s case to try, the presiding judge told the government, “I think it’s important
that we do this jointly.” Id. at 849 (quotation marks and emphasis omitted).
Eventually, the jury informed the presiding judge that they could not reach a verdict
because of one “hell bent” juror. Id. In a meeting with the presiding judge and the
government, defense counsel, unaware of the various ex parte meetings between the judge
and the government and what was discussed at those meetings, stated that the “jury should
be allowed to be hung.” Id. (quotation marks omitted). The government deferred to the
presiding judge’s discretion. Id. The presiding judge acknowledged that a unanimous vote
would not occur and decided to declare a mistrial. Id. Defense counsel requested that the
jury be polled and publicly discharged. Id. After some back and forth, the presiding judge
agreed to a public discharge, but would not allow the attorneys to poll the jurors. Id.
The defendant filed a posttrial motion for a judgment of acquittal, which was denied.
Id. A retrial was scheduled, and the defendant was found guilty of all charges. Id. After
he was convicted on retrial, the defendant came into possession of the transcript of the
sealed ex parte communications that had occurred during his first trial. Id. The defendant
9
appealed his conviction, requesting a new trial on the basis of the five ex parte meetings
and communications between the presiding judge and the government. Id. 6
The Sixth Circuit held “that these ex parte communications violated [the
defendant’s] constitutionally prescribed rights to due process, effective assistance of
counsel, and trial by an impartial judge and jury.” Id. at 850. To start, the Sixth Circuit
explained that “[a]n ex parte communication between the prosecution and the trial judge
can only be ‘justified and allowed by compelling state interest.’ ” Id., quoting United
States v Minsky, 963 F2d 870, 874 (CA 6, 1992) (additional quotation marks and citation
omitted). The government has a heavy burden to show that “the defendant was not
prejudiced when his counsel was excluded from these communications.” Id. at 851. In so
stating, the Sixth Circuit operated from the premise that the defendant was prejudiced by
those communications. The Sixth Circuit acknowledged the government’s assertion that
“its five ex parte communications were necessary to protect the existence of a wiretap in a
completely unrelated case, which was part of an ongoing criminal investigation.” Id.
(quotation marks omitted). The Sixth Circuit explained that some circumstances may
warrant “careful consideration of whether limited ex parte communications should be
allowed.” Id. (citations omitted). When this is the case, those communications should be
narrowly tailored to achieve that compelling state interest. Id. (“If ex parte
communications are to be allowed at all, they must continue no further than the extent to
6
Interestingly, the ex parte communications had not reoccurred at the defendant’s retrial.
Judge Daughtrey dissented, explaining that while she agreed with the majority’s approach,
there was no reason to give the defendant another trial when he had already received a
retrial free from ex parte communications. Barnwell, 477 F3d at 854 (Daughtrey, J.,
dissenting). The majority, however, did not believe that the second trial cured the error.
10
which they are absolutely necessary to protect the state’s compelling interest.”). The Sixth
Circuit concluded that the ex parte communications at issue did not serve a compelling
state interest and that no narrow tailoring occurred. Id.
I recognize, as does the lead opinion, that the facts of this case and Barnwell are
different. 7 The lead opinion may be correct that the facts of this case are milder than those
in Barnwell. But I see no reason why a mere factual difference precludes us from adopting
the central holding of Barnwell, that ex parte communications must be justified by a
compelling state interest and narrowly tailored when possible. Indeed, nothing in
Barnwell’s central rule was specific to or limited by the unique facts of that case—nor
should it have been, in my view. Circumstances that lead to ex parte communications will
inevitably include particularized and unique facts, though perhaps those facts will not
necessarily be as sensational as those in Barnwell. This is exactly why a broad legal
principle adaptable to different factual circumstances is an ideal solution.
Further, the Barnwell test is more practical for both defendants and prosecutors. At
no point is a defendant, under the Barnwell test, entitled to have a conviction overturned
merely because ex parte communications occurred. But a defendant proceeding under the
Barnwell test is not compelled to prove the nearly impossible—that a prosecutor changed
how it presented its case-in-chief because of ex parte communications with a judge and
7
While I recognize these factual differences, I disagree with the lead opinion’s
characterization of some of these differences. The lead opinion explains, “[U]nlike in
Barnwell, the trial judge did not cooperate with the prosecution on some matter in an effort
to further the prosecution’s interest to defendant’s detriment.” The judge in the case at
hand expressed, on more than one occasion, what she believed to be deficiencies in the
investigation and, by extension, the prosecution’s case. I cannot conclusively say that these
communications did not further the prosecution’s interest in convicting defendant.
11
that a jury’s vote was likely swayed in response to the prosecution’s alleged strategic
adjustments. And Barnwell recognizes, just as the lead opinion does, that there are some
valid reasons why an ex parte communication may occur. The Barnwell inquiry, though,
instructs a lower court to consider why those communications occurred and whether those
communications were narrowly tailored to a compelling state interest.
Moreover, I believe that Barnwell properly places the burden on the prosecution in
cases like these. Considering this case through a nonconstitutional lens, the lead opinion
imposes an incredibly high evidentiary burden on the defendant to demonstrate that the
trial judge’s and the prosecution’s misdeeds did not affect his right to a fair trial. But I
believe that the burden should instead be on the prosecution to demonstrate why its break
from the typical order was necessary. Indeed, because I agree with the Barnwell panel that
substantive ex parte communications should advance a compelling state interest, it
logically follows that the burden must be on the party who engaged in the ex parte
communication to demonstrate why it was appropriate. 8 For these reasons, I believe that
8
I acknowledge that there is a fundamental difference between this case and Barnwell. In
this case, the judge initiated contact with the county prosecutor. But in Barnwell, the
AUSA office initiated contact with the presiding judge. When a trial judge initiates
contact, it is unworkable to place the burden of proving a compelling state interest on the
judge as a non-party. Rather, this burden should still be on the prosecution. See also
Minsky, 963 F2d at 873-874 (where a presiding judge initiated ex parte conversations with
the government during the trial, the burden to provide a compelling state interest remained
with the government on appeal). Here, the prosecution’s act of engaging in this
communication and potentially altering trial strategy in response necessarily means that the
prosecution is the party who has meaningfully benefited from the ex parte communications.
It is the party who benefits from the ex parte communication on whom the burden rightfully
rests. It is also meaningful to note that the prosecution could have informed defendant of
these communications in a timely manner but did not.
12
a broad legal rule, such as the one laid down in Barnwell, is a workable standard that this
Court should adopt.
The lead opinion states that I misunderstand Barnwell and Minsky. I disagree. The
lead opinion contends that Barnwell and Minsky concern the deprivation of a defendant’s
substantive right to be present during a critical stage of the proceedings, such that neither
case supports defendant’s position here.
Again, the defendant in Barnwell was tried and convicted at a second trial, and the
ex parte communications between the presiding judge and the government occurred during
defendant’s first trial, which did not result in a conviction. No ex parte communications
occurred during the second trial at which defendant was convicted. Barnwell, 477 F3d at
853. The Sixth Circuit nonetheless questioned whether defendant received a fair second
trial before that same judge and awarded the defendant a new trial. Id. at 853-854. In other
words, the Barnwell defendant was awarded a new, third trial before a new judge on the
basis of the ex parte communications that had occurred at his first trial. I do not agree with
the lead opinion’s conclusion that defendant here has lesser substantive rights at stake than
the defendant in Barnwell, given that the ex parte communications at issue in this case
occurred during defendant’s sole trial about substantive matters associated with the trial. 9
Further, I believe that the lead opinion’s characterization of Barnwell and Minsky
fails to consider a fundamental part of the legal rule laid down in those cases. Both cases
explain that the burden to prove the lack of prejudice rests with the government. Barnwell,
9
In fact, the Barnwell panel did not specifically consider whether the defendant’s second
trial was actually marred by the ex parte communications that occurred at his first trial.
See Barnwell, 477 F3d at 852-853.
13
477 F3d at 850; Minsky, 963 F2d at 874. The defendant in each of these cases established
that ex parte communications had occurred between a presiding judge and a member of the
government during a trial proceeding. The burden was then placed on the government to
demonstrate that those communications were based on a compelling governmental interest
and that the communications were narrowly tailored. The lead opinion here “do[es] not
read Barnwell or Minsky as holding that any ex parte communication between the
prosecution and a judge that is substantive in nature violates a defendant’s right to due
process unless the government has a compelling interest for the communication.” But that
is largely what those opinions do say. See Barnwell, 477 F3d at 850 (“An ex parte
communication between the prosecution and the trial judge can only be justified and
allowed by compelling state interest.”) (emphasis added; quotation marks and citations
omitted); id. at 851 (“If ex parte communications are to be allowed at all, they must
continue no further than the extent to which they are absolutely necessary to protect the
state’s compelling interest.”) (emphasis added); Minsky, 963 F2d at 874 (“Although there
are circumstances where an ex parte communication might be overlooked, the burden of
proving lack of prejudice is on the [government], and it is a heavy one.”) (emphasis added;
quotation marks and citations omitted); id. (clarifying that ex parte proceedings “can only
be justified and allowed by compelling state interests”) (emphasis added; quotation marks
and citation omitted). These cases use mandatory language to say that ex parte
communications can only occur to further a state’s compelling interest. In other words, I
believe that these cases explain that where an ex parte communication between the
government and a presiding judge occurs, a presumption of prejudice generally arises. In
distinguishing Barnwell and Minsky from this case, the lead opinion considers defendant’s
14
due-process claims in reverse. The lead opinion appears to recognize that Barnwell and
Minsky place the burden on the prosecution to establish that ex parte communications were
justified. But the lead opinion seems to conclude that here, the prosecution need not answer
for its ex parte communications because defendant cannot establish prejudice.
To support this approach, the lead opinion also concludes that the e-mail exchange
with the county prosecutor was not a critical stage of the proceedings. In my view, in
attempting to look at the e-mail exchange in isolation from the rest of the trial proceedings,
the lead opinion walks a tightrope too thin to support its own weight. As the lead opinion
itself explained, the e-mail exchange occurred during live witness testimony in the midst
of defendant’s trial, and the e-mails involved substantive matters associated with
defendant’s trial. See People v King, 512 Mich 1, 11; 999 NW2d 670 (2023) (explaining
that trial is a critical stage of criminal proceedings). The lead opinion has now taken the
view that a substantive ex parte communication about defendant’s trial that occurred during
defendant’s trial somehow does not occur during a critical stage of that trial. I do not
believe that the lead opinion has supported its conclusion that the e-mail exchange here did
not occur at a critical stage of the proceedings.
With this background in mind, my next consideration is whether defendant here was
deprived of due process under the Barnwell test. If so, he is entitled to a new trial. At this
juncture, I reiterate once more that not all ex parte communications result in a constitutional
violation. Barnwell itself recognized that there are circumstances in which such
communications might be necessary. Barnwell, 477 F3d at 850, citing Minsky, 963 F2d at
874. See generally United States v Gagnon, 470 US 522, 526; 105 S Ct 1482; 84 L Ed 2d
486 (1985) (explaining that not every ex parte conversation between a trial judge and a
15
juror violates the Constitution); Michigan Code of Judicial Conduct, Canon 3(A)(4)
(allowing ex parte communications for administrative purposes). 10
In this case, the trial judge communicated with the county prosecutor during the
trial. In each communication, the trial judge highlighted what she believed to be flaws in
the criminal investigation and, by extension, the presentation of the prosecution’s case-in-
chief. In turn, these communications could have altered how the prosecution decided to
present those issues to the jury. At oral argument, many of my colleagues questioned the
extent to which the trial judge here exhibited bias and to what extent the trial judge’s
communications with the county prosecutor could have reasonably altered the
prosecution’s presentation of the case, such that defendant could establish he was denied
due process and entitled to relief. But this is a futile, if not impossible, task.
Again, this is a task neither the courts nor aggrieved defendants need complete under
the Barnwell test. Instead, under the Barnwell test, prejudice is presumed when a judge
engages in substantive ex parte communications with a prosecutor. See Barnwell, 477 F3d
at 851 (“The Government bears a heavy burden in showing that the defendant was not
prejudiced when his counsel was excluded from these communications.”); Minsky, 963 F2d
at 874 (“Although there are circumstances where an ex parte communication might be
overlooked, the burden of proving lack of prejudice is on the government, and it is a heavy
10
In Michigan, a judge can engage in some ex parte discussions with a party if they relate
to administrative matters and all parties are placed on notice that the communications
occurred. But like the lead opinion, I do not believe that the discussions here involved
administrative matters, and, moreover, defendant was never apprised of the
communications while he would have had an opportunity to object or seek any relief.
Because no valid reason for these ex parte communications is apparent from my review of
the case, I would start with the presumption that defendant was prejudiced.
16
one.”) (quotation marks, alterations, and citations omitted). I also find support in how other
courts have imposed presumptions of prejudice in cases involving ex parte
communications. See Remmer v United States, 347 US 227, 229; 74 S Ct 450; 98 L Ed
654 (1954) (recognizing that in some circumstances a rebuttal presumption arises when
communication occurs with a juror); France, 436 Mich at 143 (“A substantive
communication carries a presumption of prejudice in favor of the aggrieved party
regardless of whether an objection is raised.”); United States v Mesteth, 528 F2d 333, 335
(CA 8, 1976) (“This court has recognized that a presumption of prejudice arises from a
court’s instructing the jury outside the presence of the defendant, but that the presumption
‘in some instances may be overcome by evidence giving a clear indication of lack of
prejudice.’ ”) (citation omitted).
Having adopted the Barnwell test, I would start from the presumption that the ex
parte communications prejudiced defendant. The burden would thus be on the prosecution
to establish (1) whether the ex parte communication between the county prosecutor and the
trial judge was justified by a compelling state interest; and if so, (2) whether those
communications were narrowly tailored to advance that compelling state interest. If the
prosecution could not satisfactorily answer these questions, defendant would be entitled to
a new trial.
III. CONCLUSION
This case presents an opportunity to opine on the proper legal inquiry when a judge
engages in ex parte communications with a prosecutor. Categorizing these errors as
nonconstitutional, as the lead opinion does, neglects to consider the full range of effects
17
that these communications may impose on the right to a fair trial and how ex parte
communications with one party have been persuasively analyzed by other courts. To the
extent that the lead opinion performs a constitutional analysis, it falls short by overstating
the differences between this case and Barnwell. I would instead take this opportunity to
adopt the Barnwell test. I believe that the legal conclusions reached by the Sixth Circuit in
Barnwell comport with decisions this Court has made under factually similar
circumstances. In my view, Barnwell is a natural extension of this Court’s opinion in
France. Rather than proceed in the manner the lead opinion does, I would take this
opportunity to adopt a test that more specifically addresses issues posed by ex parte
communications between a judge and a prosecutor. I am satisfied that the Barnwell test is
relevant, is practical, and adequately captures the historical concept that ex parte
communications are tolerable only under rare circumstances.
However, the parties have not had an opportunity to fully brief an argument under
this test. 11 Accordingly, I would adopt the Barnwell test and remand this case to the Court
of Appeals, permitting the parties an opportunity to fully brief this issue.
For these reasons, I respectfully dissent.
Richard H. Bernstein
11
Defendant, in his supplemental brief, summarily stated that if this Court does not adopt
the Liljeberg factors, this Court could look to the Barnwell test. The prosecution’s brief
did not address Barnwell.
18