NOT RECOMMENDED FOR PUBLICATION File Name: 24a0198n.06
Case No. 23-1881 FILED UNITED STATES COURT OF APPEALS May 03, 2024 FOR THE SIXTH CIRCUIT KELLY L. STEPHENS, Clerk
) JEFFREY HAROLD STEFANSKI, ) Petitioner - Appellant, ) ) ON APPEAL FROM THE UNITED v. ) STATES DISTRICT COURT FOR THE ) EASTERN DISTRICT OF MICHIGAN ADAM DOUGLAS, Warden, ) Respondent - Appellee. ) OPINION )
Before: GIBBONS, MCKEAGUE, and STRANCH, Circuit Judges.
JULIA SMITH GIBBONS, Circuit Judge. Jeffrey Harold Stefanski was convicted in a
Michigan state court of three counts of criminal sexual conduct. He appealed his conviction in
state court, raising claims of ineffective assistance of counsel based on (1) his counsel’s failure to
introduce evidence that would have supported Stefanski’s theory that the victim made up the
sexual assaults to curry favor in a separate court proceeding, and (2) his counsel’s failure to
successfully object to the admission of hearsay testimony by the victim’s mother. After the
Michigan courts affirmed on direct review, Stefanski filed a habeas petition in federal court. The
district court denied Stefanski habeas relief after determining that, even if Stefanski’s counsel was
deficient, Stefanski could not establish that he was prejudiced by his counsel’s alleged errors. For
the following reasons, we affirm. No. 23-1881, Stefanski v. Douglas
I.
Stefanski’s convictions stem from his sexual assault of Phil Lounds.1 Stefanski—forty-six
years old at the time of trial—owned a home and cabin in Northern Michigan, where he allegedly
hosted sexually-charged parties in which he allowed teenage boys to drink alcohol. People
v. Stefanski, No. 334510, 2018 WL 5276411, at *1 (Mich. Ct. App. Oct. 23, 2018). Lounds—then
eighteen years old—attended some of these parties with friends. Lounds alleged that Stefanski
twice inserted his penis into Lounds’s mouth when Lounds was too intoxicated to stop him. Id.
On another occasion, Stefanski attempted to initiate anal sex with Lounds while Lounds was
“nearly unconscious.” Id. This time, Lounds was able to roll over and prevent penetration. Id.
Lounds first reported these sexual assaults to a lawyer, Jennifer France, who represented
him in a pending probation violation proceeding. Id. at *2. Stefanski was then charged with two
counts of criminal sexual conduct in the first degree and one count of criminal sexual conduct in
the second degree. Stefanski’s first trial resulted in a mistrial, but he was convicted in the second.
At trial, Lounds testified that he reported Stefanski’s misconduct while awaiting sentencing for his
parole violation after seeing on the news that Stefanski was charged with a different crime
involving sexual misconduct. While Lounds asserted that he did not expect to get anything out of
reporting Stefanski’s crimes, Stefanski’s theory of defense was that Lounds falsified the sexual
assaults to get a more lenient sentence at his probation violation sentencing. To this end,
Stefanski’s trial counsel underscored the timing and inconsistencies of Lounds’s first reports of the
sexual assaults—noting that Lounds first told his lawyer as he faced a pending violation
proceeding, that Lounds spent significant time at Stefanski’s home after the assaults allegedly
1 We recite the facts largely as portrayed by the Michigan Court of Appeals. This recitation is presumed to be correct. See Wagner v. Smith, 581 F.3d 410, 413 (6th Cir. 2009) (citing 28 U.S.C. § 2254(e)(1)). Stefanski does not dispute the Michigan court’s discussion of the facts on appeal. -2- No. 23-1881, Stefanski v. Douglas
occurred, and introducing testimony conflicting with Lounds’s accounts of Stefanski’s
sexually-charged parties. Stefanski also took the stand and refuted Lounds’s and others’ accounts
that he regularly served teenagers excessive alcohol and engaged in sexual behavior at gatherings
he hosted; he further adamantly denied Lounds’s allegations of sexual assault.
The jury convicted Stefanski on all three counts. Ultimately, the court sentenced Stefanski
to 11 to 40 years’ imprisonment on counts 1 and 2 to run concurrently, and 2 to 15 years on count
3—also concurrent. On direct appeal and collateral review, Stefanski has maintained that his
counsel performed deficiently in two instances during trial, which amounted to ineffective
assistance of counsel and resulted in convictions in violation of Stefanski’s constitutional rights.
First, Stefanski challenges his counsel’s failure to introduce Lounds’s probation records to
conclusively demonstrate that Lounds faced a pending probation violation proceeding when he
first disclosed Stefanski’s sexual assault to France. This, Stefanski argues, was key to establishing
Lounds’s motive to fabricate the assaults. Although Lounds readily admitted in his testimony that
he faced sentencing for a probation violation when he first reported the sexual assault, France
equivocated. France agreed that Lounds was on probation when he disclosed the assaults and that
she later used his reporting in a sentencing memorandum, but she denied that Lounds faced charges
of new crimes when he first reported. Further, she testified that “nothing was pending” at the time
of his initial report. DE 5-15, Trial Tr. Vol. II, Page ID 944. Lounds’s probation records, which
were not introduced at trial, confirm that Lounds was facing a pending parole violation when he
first disclosed Stefanski’s assaults to France. Stefanski, 2018 WL 5276411, at *2 n.2. Lounds had
also garnered a second parole violation a few days after his disclosure. France testified that
Lounds’s upcoming parole hearing, and her later use of his disclosure in a sentencing
memorandum, referred to this later violation that occurred after his disclosure, and not any prior
-3- No. 23-1881, Stefanski v. Douglas
violation. This underscored her (incorrect) testimony that Lounds was not facing a pending
violation proceeding when he first reported Stefanski’s crimes. Because Stefanski’s theory of
defense—that Lounds had a motive to falsify the allegations—depended on Lounds facing jail
time at the time of his initial disclosure, Stefanski argues that his counsel’s failure to impeach
France with Lounds’s probation records to reaffirm that he was in fact facing a probation violation
greatly undermined his defense. He points to the government’s use of France’s testimony in
closing arguments to say that Lounds “was not in trouble” when he first reported the assault as
indicative of the harm done. DE 5-16, Trial Tr. Vol. III, Page ID 1118, 1151.
Second, Stefanski charges his counsel as ineffective for failing to successfully object to
hearsay testimony from Lounds’s mother. Lounds’s mother briefly testified to the contents of a
phone conversation she had with Lounds while he was in jail, presumably for the probation
violation. She recalled Lounds crying and telling her that Stefanski had done something sexual to
him. Stefanski’s counsel lodged a hearsay objection, but the court ultimately overruled the
objection and agreed with the prosecution that the testimony could come in as a prior consistent
statement under the Michigan Rules of Evidence. Stefanski argues that his counsel erred in failing
to inform the judge that the testimony was not in fact a prior consistent statement because Lounds
made the statement after he had developed the motive to fabricate the assault.
After the trial court denied Stefanski’s motion for a new trial on these grounds, Stefanski
unsuccessfully raised both issues on direct appeal in state court. Stefanski, 2018 WL 5276411, at
*2–5. The Michigan Court of Appeals rejected Stefanski’s first assignment of error after finding
Stefanski’s counsel not deficient in his failure to introduce Lounds’s probation records.
Specifically, the court noted that Lounds “himself readily admitted that he was facing a probation
violation when he disclosed the sexual assault. Therefore, the probation documents would have
-4- No. 23-1881, Stefanski v. Douglas
demonstrated that the victim testified truthfully, bolstering his credibility.” Id. at *2. Further, the
court highlighted that France’s testimony about the pending proceeding was ambiguous,
thus reducing the need to impeach her, as she also testified that she did attempt to obtain a reduced
sentence for Lounds based on his cooperation in the sexual assault case against Stefanski.
Id. Even if counsel did unreasonably err, the court found that Stefanski failed to establish
prejudice, as Lounds’s credibility was supported by more than his own testimony; other witnesses
corroborated aspects of Lounds’s account, such as the sexual nature of the parties and Stefanski’s
inappropriate behavior toward Lounds and others. Id. at *3. Likewise, the court rejected
Stefanski’s second allegation of error: the missed hearsay objection. The court did agree with
Stefanski that the testimony was inadmissible hearsay. Id. at *5. But even assuming counsel
performed deficiently by failing to raise the correct hearsay objection, the court concluded that
Stefanski again failed to show prejudice. The court found Lounds’s mother’s testimony vague and
unlikely to carry much weight in light of the other testimony from witnesses detailing more specific
aspects of Stefanski’s parties where the assaults took place. Id. Furthermore, since the statement
was made after Lounds was imprisoned for the probation violation, it “actually bolstered”
Stefanski’s motive-to-falsify theory. Id.
Stefanski then reasserted these claims in state and federal court. First, he applied for leave
to appeal to the Michigan Supreme Court, which was denied. People v. Stefanski, 924 N.W.2d
552 (Mich. 2019) (mem). Then, Stefanski filed a habeas petition in district court. After finding
the Michigan Court of Appeals’s decision properly within 28 U.S.C. § 2254(d)’s realm, the district
court rejected Stefanski’s petition but granted a certificate of appealability. Stefanski appealed.
-5- No. 23-1881, Stefanski v. Douglas
II.
When reviewing the denial of habeas corpus relief, we review the district court’s legal
conclusions de novo and its findings of fact for clear error. Rayner v. Mills, 685 F.3d 631, 636
(6th Cir. 2012). Our review of the state court’s determinations, however, is governed by the
Antiterrorism and Effective Death Penalty Act (“AEDPA”), 28 U.S.C. § 2254(d). Fleming
v. Metrish, 556 F.3d 520, 524 (6th Cir. 2009) (citing Ivory v. Jackson, 509 F.3d 284, 291 (6th Cir.
2007)). AEDPA provides that:
(d) An application for a writ of habeas corpus on behalf of a person in custody pursuant to the judgment of a State court shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim—
(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or
(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.
28 U.S.C. § 2254(d).
A state court’s decision is “contrary to” clearly established federal law “if the state court
applies a rule different from the governing law set forth in [Supreme Court] cases, or if it decides
a case differently than [the Supreme Court has] done on a set of materially indistinguishable facts.”
Lang v. Bobby, 889 F.3d 803, 810 (6th Cir. 2018) (quoting Bell v. Cone, 535 U.S. 685, 694 (2002)).
Alternatively, a state court’s decision denying habeas relief is an “unreasonable application” of
clearly established federal law if the court identifies the correct governing legal principle but
applies it in an “objectively unreasonable” way to the facts at hand. Wiggins v. Smith, 539 U.S.
510, 520–21 (2003) (quoting Williams v. Taylor, 529 U.S. 362, 409, 413 (2000)). The state court’s
-6- No. 23-1881, Stefanski v. Douglas
determination of a claim is insulated if “fairminded jurists could disagree” on the correctness of
its decision. Harrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,
541 U.S. 652, 664 (2004)). In applying these standards, we look to the last reasoned state court
decision. Lang, 889 F.3d at 813. Here, that is the decision of the Michigan Court of Appeals.
III.
A claimant establishes ineffective assistance of counsel in violation of his Sixth
Amendment right by showing that his counsel’s conduct fell below “an objective standard of
reasonableness” and that a “reasonable probability” exists that “but for counsel’s unprofessional
errors, the result of the proceeding would have been different.” Strickland v. Washington, 466 U.S.
668, 688, 694 (1984). In determining whether counsel’s conduct was objectively unreasonable,
courts “must apply a ‘strong presumption’ that counsel’s representation was within the ‘wide
range’ of reasonable professional assistance.” Richter, 562 U.S. at 104 (quoting Strickland,
466 U.S. at 689). And a “reasonable probability” of a different outcome “is a probability sufficient
to undermine confidence in the outcome.” Id. (quoting Strickland, 466 U.S. at 694).
Because Stefanski seeks habeas corpus relief, AEDPA adds another layer to Strickland’s
two-part test. Id. at 105. As the Supreme Court recognized, “[t]he standards created by Strickland
and § 2254(d) are both ‘highly deferential,’ and when the two apply in tandem, review is ‘doubly’
so.” Id. (internal citations omitted). “When § 2254(d) applies, the question is not whether
counsel’s actions were reasonable. The question is whether there is any reasonable argument that
counsel satisfied Strickland’s deferential standard.” Id. Stefanski argues that the Michigan
Court of Appeals assessed Stefanski’s ineffective assistance claims in a manner that was both
(1) contrary to clearly established law, and (2) an unreasonable application of federal law. We
discuss both arguments in turn.
-7- No. 23-1881, Stefanski v. Douglas
1.
The Michigan Court of Appeals and the district court found that counsel’s failure to
introduce Lounds’s probation records did not support a claim for ineffective assistance of counsel.
We affirm. First, Stefanski cannot surpass AEDPA’s restriction of our consideration of the state
court’s assessment of his attorney’s performance. And even if his attorney was deficient, the state
court’s assessment of prejudice is also entitled to AEDPA deference.
A. Performance
Stefanski first argues that the Michigan Court of Appeals unreasonably applied
Strickland’s performance prong in finding his counsel’s failure to introduce Lounds’s probation
records not deficient, thus making AEDPA’s prohibition on relief inappropriate. According to
Stefanski, “no reasonable, fairminded jurist could subscribe” to the Michigan court’s analysis.
CA6 R. 5, Petitioner Br., at 34–35 (citing Richter, 562 U.S. at 101). Counsel’s mistake was
unreasonable, in Stefanski’s eyes, because using Lounds’s probation records to contextualize
France’s testimony would have helped Stefanski’s motive-to-falsify theory more than it would
have bolstered Lounds’s credibility. In light of this cost-benefit analysis, and lacking any reason
not to introduce the records, no fairminded jurist could find defense counsel’s failure to do so
within the realm of reasonable representation.
True, Stefanski’s lawyers conceded that they possessed documentation showing that
Lounds had a pending probation violation proceeding when he reported the assaults to France.
And the government later capitalized on the timing ambiguity in closing arguments, invoking
France’s testimony to say that Lounds was not in any trouble and thus lacked a reason to fabricate
the assaults when he first reported them. The better course may well have been introducing
Lounds’s records to settle any remaining confusion on the matter. But the state court addressed
-8- No. 23-1881, Stefanski v. Douglas
these arguments and found Stefanski’s counsel’s conduct within the wide realm of reasonable
representation. See Stefanski, 2018 WL 5276411, at *2. In doing so, the state court identified
“reasonable argument[s] that counsel satisfied Strickland’s deferential standard.” Richter,
562 U.S. at 105. Even if the state court’s reasoning was “incorrect or erroneous,” Wiggins,
539 U.S. at 520, a petitioner alleging an unreasonable application of clearly established federal
law “must show far more than that the state court’s decision was ‘merely wrong’ or ‘even clear
error.’” Shinn v. Kayer, 592 U.S. 111, 118 (2020) (per curiam) (quoting Virginia v. LeBlanc,
582 U.S. 91, 94 (2017) (per curiam)). Stefanski has not done so, and AEDPA accordingly
prohibits relief.
The state court found that defense counsel’s failure to impeach France with the records
should be viewed in light of the other evidence introduced, namely, Lounds’s admission that he
faced a probation violation when he first disclosed the assault. Stefanski, 2018 WL 5276411, at
*2. Strickland supports this. See Strickland, 466 U.S. at 690 (considering the reasonableness of
counsel’s conduct “in light of all the circumstances”). Stefanski himself painted this case as a
credibility contest between Lounds and himself. The state court found that impeaching France
further could have incidentally bolstered Lounds’s credibility by confirming that portion of his
testimony. Stefanski, 2018 WL 5276411, at *2. The court also highlighted the inconsistencies
inherent in France’s testimony, as she admitted to using Lounds’s sexual assault report in a
sentencing memorandum—making impeachment of her testimony less essential. Id.
The state court’s weighing of the risk of crediting Lounds’s testimony may have been
incorrect, but its determination that Stefanski’s counsel fell within Strickland’s broad realm of
reasonableness is not “objectively unreasonable.” Williams, 529 U.S. at 409. As the district court
recognized, France’s testimony was unclear at times as to whether she was referring to Lounds not
-9- No. 23-1881, Stefanski v. Douglas
having an outstanding criminal charge (which was true) or not having a probation violation (which
was not true). It only became clear at closing argument that the government used France’s
testimony to establish that there was no pending probation violation at all. Although the preferable
way to clear up the confusion would have been to introduce Lounds’s probation records during
France’s testimony, Stefanski’s attorney instead raised France’s prior testimony detailing
Lounds’s pending probation violation, although that proved unsuccessful.
The circumstances of this case thus differ from those presented in Peoples v. Lafler,
734 F.3d 503 (6th Cir. 2013), in which this court found that an ineffective assistance claim based
on counsel’s failure to present impeachment evidence warranted habeas relief. Id. at 512. In
Lafler, defense counsel failed to impeach known false testimony from the only two witnesses who
tied the defendant to the crime. See id. at 512–13. The court found that the defense counsel thus
failed to elicit key testimony, which was the only way for the defendant to prove his theory of the
case. Id. at 513. Here, in contrast, the information Stefanski desired was already in the record
through Lounds, who testified the day before France and repeatedly affirmed that he faced an
upcoming sentencing for a probation violation when he first disclosed the sexual assaults to France.
And Stefanski’s counsel did attempt to draw out France’s use of Lounds’s reporting to obtain a
more lenient sentence in a proceeding, comporting with Stefanski’s general theory that Lounds
fabricated the story to benefit himself in legal proceedings. Although the probation record could
have easily been introduced, weighing in favor of deficient performance, see Lafler, 734 F.3d at
514, the misstep does not rise to the magnitude warranting habeas relief. At bottom, although
Stefanski’s counsel’s conduct was certainly not the best course of action, we are not faced “with a
record that cannot, under any reasonable interpretation of the controlling legal standard, support”
- 10 - No. 23-1881, Stefanski v. Douglas
the state court’s finding that Stefanski’s counsel was not deficient. Panetti v. Quarterman,
551 U.S. 930, 953 (2007); see also Knowles v. Mirzayance, 556 U.S. 111, 126–27 (2009).
B. Prejudice
If Stefanski’s counsel was deficient, such deficiency gives rise to a constitutional violation
only upon a showing of a reasonable probability that, but for counsel’s deficiency, the outcome of
the proceeding would have been different. Strickland, 466 U.S. at 694. Stefanski attacks the state
court’s conclusion that he failed to show prejudice both as contrary to clearly established federal
law and as an unreasonable application of federal law. Taking each in turn, we disagree.
Contrary to Clearly Established Federal Law. Where a state court applies a legal rule
contrary to governing Supreme Court law, a federal court on review is “unconstrained by
§ 2254(d)(1)” and can review the petitioner’s claim de novo. Fulcher v. Motley, 444 F.3d 791,
799 (6th Cir. 2006) (quoting Williams, 529 U.S. at 406). A paradigmatic example would be a state
court stating that an ineffective assistance of counsel claim requires the petitioner to prove by a
preponderance of the evidence—rather than Strickland’s reasonable probability standard—that the
result of his proceeding would have been different absent counsel’s errors. Williams, 529 U.S. at
405–06. That said, we must avoid a “readiness to attribute error” to state court decisions and
instead view them “with the presumption that state courts know and follow the law.” Woodford
v. Visciotti, 537 U.S. 19, 24 (2002) (per curiam).
In line with this presumption, an occasional imprecise or incorrect articulation of federal
law does not run afoul of AEDPA if the state court otherwise properly discussed the governing
law. See id. at 23–24 (imprecise recitation of prejudice prong does not automatically render state
court decision contrary to clearly established law); Holland v. Jackson, 542 U.S. 649, 654–55
(2004) (per curiam) (finding articulation of Strickland not contrary to clearly established law even
- 11 - No. 23-1881, Stefanski v. Douglas
though state court could be read to implement a preponderance of the evidence standard for
prejudice prong but such reading would “needlessly create internal inconsistency in the opinion,”
and the state court otherwise correctly articulated Strickland’s test); Johnson v. Genovese,
924 F.3d 929, 937–38 (6th Cir. 2019) (not contrary to federal law where state court articulated
Strickland standard correctly multiple times and only articulated an incorrect standard once);
see also Daniel v. Curtin, 499 F. App’x 400, 411 n.5 (6th Cir. 2012) (not contrary to federal
law where court correctly articulated Strickland standard but then twice referenced
outcome-determinativeness and the parties did not raise the issue). But see Vasquez v. Bradshaw,
345 F. App’x 104, 111–12 (6th Cir. 2009) (contrary to clearly established law when the state court
repeatedly articulated and applied a but-for test instead of Strickland’s reasonable probability test);
Spaulding v. Larson, 704 F. App’x 475, 480 (6th Cir. 2017) (contrary to clearly established law
where “the Michigan Court of Appeals never stated the Strickland standard correctly, and thrice
stated it incorrectly”); Magana v. Hofbauer, 263 F.3d 542, 550 (6th Cir. 2001) (contrary to clearly
established law where the state court clearly articulated and applied a stricter prejudice standard).
Stefanski argues that AEDPA deference is unwarranted here because the state court
improperly jettisoned Strickland and required Stefanski to show that his counsel’s error was
“outcome determinative.” Stefanski, 2018 WL 5276411, at *3.2 Stefanski is correct that the state
court’s reference to outcome determinativeness could be read as holding Stefanski to a higher
burden than Strickland requires. But in line with the Supreme Court’s command that we view
2 In its discussion of whether the failure to introduce Lounds’s probation records prejudiced Stefanski, the Michigan Court of Appeals first cited Strickland’s reasonable probability standard. Stefanski, 2018 WL 5276411, at *3 (citing analogous standard under Michigan law). After discussing the other evidence that supported Stefanski’s conviction, the court ended the paragraph with its finding that “[t]hus, the jury considered other evidence that corroborated the victim’s testimony, and the contradiction about the timing of the victim’s probation violation proceedings was not outcome determinative.” Id. Stefanski takes issue with this last sentence. - 12 - No. 23-1881, Stefanski v. Douglas
state court opinions with a presumption that they consistently and correctly apply the law,
see Holland, 542 U.S. at 654–55, the state court’s opinion is more properly read as following
Strickland, with the last sentence inartfully crafted. At the outset, the state court thrice articulated
Strickland’s prejudice prong—with its reasonable probability requirement—correctly, including
once in the very same paragraph in which the state court wrote the offending sentence.
See Stefanski, 2018 WL 5276411, at *1, 3, 4. And the court’s reference to outcome
determinativeness appears in the context of responding to Stefanski’s claim that the counsel’s
failure was “fatal to his defense,” suggesting that the court’s phrasing may have been a response
to his argument rather than the application of a stricter prejudice test. Id. at *3. In light of the
court’s repeated correct articulation of the law, we decline to “needlessly create internal
inconsistency” by charging the court with applying the incorrect test based on this stray reference.
Holland, 542 U.S. at 654; see also Genovese, 924 F.3d at 937–38, 937 n.2 (finding case with
similar error more like Holland than Vasquez).
Unreasonable Application. Stefanski next contends that even if the state court’s prejudice
determination was not contrary to clearly established federal law, AEDPA is inapplicable because
it nonetheless constitutes an unreasonable application of federal law. We disagree.
Stefanski bases this argument on the proposition that the state court credited the
government’s witnesses over Stefanski’s, thus improperly weighing the evidence and usurping the
role of the jury. But Stefanski misinterprets the state court’s opinion and AEDPA’s restrictions
on our review. Determining whether a reasonable probability of a different outcome exists requires
reviewing courts to take stock of the strengths and weaknesses of the surrounding case.
See Richter, 562 U.S. at 112–13 (discussing the strength of the other evidence of guilt); Kayer,
592 U.S. at 121–22 (applying similar analysis in the context of mitigation). And here, the state
- 13 - No. 23-1881, Stefanski v. Douglas
court reasonably concluded that, in light of Lounds’s repeated testimony that he faced a pending
probation violation at the time of his initial report, as well as testimony by other men corroborating
aspects of his story, Stefanski failed to demonstrate a reasonable probability that he would not
have been convicted had the probation records been introduced.3 Stefanski, 2018 WL 5276411, at
*3. The import of the state court’s analysis was not to make a credibility determination favoring
the government’s witnesses over Stefanski’s testimony and witnesses, but to demonstrate that
Lounds’s alleged motive to fabricate was not the key to the case, as the case turned on not only
Lounds’s testimony, but the credibility of other witnesses as well.4 The court could have
reasonably found that the case did not turn mainly on the motive theory, as the testimony
concerning the sexual nature of Stefanski’s parties differed so greatly between the government and
defense’s witnesses—with the government’s witnesses describing Stefanski hosting parties
featuring heavy intoxication and overt sexual activities while the defense’s witnesses described
more casual gatherings along the lines of card games and soda pop or light drinking—the jury
simply had to choose to believe one side over the other. Stefanski’s disagreement with the state
3 We also note that while Stefanski attributes the change in verdict across the first and second trial to France’s different testimony, Stefanski recognizes that two of the government’s witnesses who corroborated aspects of Lounds’s accounts of sexually-charged parties had not testified in the first trial. The addition of these two witnesses to corroborate Lounds’s account and refute Stefanski’s narrative of events undoubtedly strengthened the government’s case. 4 Further, Ramonez v. Berghuis, 490 F.3d 482 (6th Cir. 2007), a case cited by Stefanski, speaks largely to a different issue than presented here. The court in Ramonez rejected the notion that the state court’s credibility determinations were entitled to deference under § 2254(e)(1)—a portion of AEDPA not implicated in the case before us. See 490 F.3d at 491 (finding determinations within the prejudice prong to be a mixed issue of law and fact and thus not subject to § 2254(e)(1)’s presumption). Although it also deemed the state court’s determination that the defendant was not prejudiced unreasonable, it did so under circumstances different than here: the defense counsel failed to conduct a reasonable investigation to learn that witnesses could have disputed a key element of the prosecution’s case on which the jury had indicated it was deadlocked. Id. at 489. - 14 - No. 23-1881, Stefanski v. Douglas
court’s conclusion falls far short of establishing that no fairminded jurist could agree that he failed
to establish prejudice. See Richter, 562 U.S. at 102.
As the Supreme Court has noted, a reasonable probability requires more than a
“‘conceivable[]’ likelihood of a different result.” Kayer, 592 U.S. at 118 (internal citation
omitted). It is thus not enough for Stefanski to show that the introduction of Lounds’s probation
records might have generated a “somewhat stronger defense and the jury conceivably might have
reached a different result.” Spaulding, 704 F. App’x at 483. We have held that where evidence
that counsel failed to introduce at trial “would have been simply cumulative to the evidence
actually presented to the jury,” this failure “does not rise to the level of a constitutional violation.”
Getsy v. Mitchell, 495 F.3d 295, 313 (6th Cir. 2007) (en banc) (quoting Broom v. Mitchell,
441 F.3d 392, 410 (6th Cir. 2006)). This circuit has recognized that in cases of sexual assault,
which often lack physical evidence, prejudice may be found more readily. See Vasquez, 345 F.
App’x at 119 (quoting Hodge v. Hurley, 426 F.3d 368, 386 (6th Cir. 2005)). But a petitioner
alleging prejudice still must point to evidence that “differ[s] in a substantial way—in strength and
subject matter—from the evidence actually presented” at trial. Hill v. Mitchell, 400 F.3d 308,
319 (6th Cir. 2005); see Martin v. Haas, 731 F. App’x 443, 458–59 (6th Cir. 2018) (deeming
failure to introduce impeachment evidence of rape victim that would have supported her motive
to fabricate charges not prejudicial because it was only “relatively more compelling” than the
theory of motive presented at trial); Hartman v. Bagley, 492 F.3d 347, 361 (6th Cir. 2007) (finding
no prejudice where counsel failed to expand on or corroborate information already in the record).
Stefanski’s case does not resemble those in which we have found prejudice. Cf. Lewis
v. Wilkinson, 307 F.3d 413, 422 (6th Cir. 2002) (granting habeas petition in rape case due a
Confrontation Clause violation where the state trial court wrongfully excluded statements from
- 15 - No. 23-1881, Stefanski v. Douglas
victim’s diary indicating that the sexual conduct was consensual, and the victim had other motives
for pursuing prosecution). In Lewis, the failure to introduce excerpts from the victim’s diary
strongly suggesting that she fabricated the sexual assault charges was not harmless in part because
the excerpts went to a “different type of motive than that implied by the other evidence”
introduced. Id. And in Vasquez, the defense counsel failed to introduce key testimony from
multiple witnesses strongly calling into question the child-victim’s allegations in a manner not
otherwise conveyed to the jury. 345 F. App’x at 119–21. Here, in contrast, the introduction of the
probation records would have merely corroborated the information Stefanski’s counsel had elicited
from Lounds himself—that he faced probation violation proceedings when he first reported the
sexual assault. Lounds did not mention the probation proceedings just one time in passing—he
noted three times his understanding that he faced an upcoming sentencing when he first reported
the sexual assaults to France. Even if, as Stefanski alleges, the jury may have believed, after
France’s testimony, that Lounds was mistaken, Lounds’s testimony at least established that he
believed he faced an upcoming sentencing for such a violation when he first reported the assaults
and thus still supported the motive-to-fabricate theory.
The government’s use of France’s testimony to suggest that Lounds faced no probation
violation proceeding is certainly concerning, but it did not render the state court’s determination
beyond reasonable debate, nor did it establish prejudice. Based on Lounds’s repeated affirmations
that he faced a pending sentencing, it appears illogical and unlikely that the jury affirmatively
disbelieved Lounds’s statements that he faced sentencing for a pending violation at the time of his
initial report, but simultaneously believed Lounds’s allegations—over testimony by Stefanski and
others to the contrary—that Stefanski sexually assaulted Lounds on three occasions. Because the
record already contained much of the information the probation record would have confirmed,
- 16 - No. 23-1881, Stefanski v. Douglas
fairminded jurists could disagree as to whether the failure to introduce the probation records tipped
the scales. Because “prejudice is at best debatable, . . . we are therefore not empowered on federal
habeas review to disturb the Michigan Court of Appeals’s rejection” of Stefanski’s ineffective
assistance claim. Martin, 731 F. App’x at 459; see 28 U.S.C. § 2254(d).
2.
Stefanski also charges ineffective assistance for his counsel’s failure to lodge the correct
objection to hearsay testimony from Lounds’s mother. For the following reasons, we disagree.
The parties first disagree about the import of the state court’s discussion of performance,
and thus which standard of review applies to this prong.5 Even assuming that de novo review
applies to this prong because it is unclear whether the state court assessed performance on the
merits, see Rayner, 685 F.3d at 638, we do not find counsel’s performance deficient.
Lounds’s mother testified that Lounds called her from jail and disclosed that Stefanski “did
something [sexual] to him.” DE 5-14, Trial Tr. Vol. I, Page ID 921. Defense counsel lodged a
timely objection on the basis of hearsay, which the court initially sustained. After the prosecution
argued that the statement was admissible as a prior consistent statement, the court reversed course
and admitted the testimony. The Michigan Court of Appeals found that because Lounds made the
statement after the motive to fabricate arose, the statement was not in fact admissible as a prior
5 In addition to arguing that AEDPA deference applies, the government also argues that this is a matter of state law and thus unreviewable by federal courts. This argument is misguided, as the failure to move to suppress evidence admitted pursuant to state evidentiary law is cognizable as an ineffective assistance of counsel claim. See, e.g., Hewitt-El v. Burgess, 53 F.4th 969, 980 (6th Cir. 2022) (finding counsel’s impeachment of his own client with his five previous felony convictions, as well as his failure to attempt to exclude such convictions, unreasonable). - 17 - No. 23-1881, Stefanski v. Douglas
consistent statement. See Stefanski, 2018 WL 5276411, at *5. Stefanski thus faults his counsel
for not following up with a more specific objection to rebut the prosecutor’s (incorrect) contention.
Stefanski highlights the fact that his counsel lacked a strategic reason for failing to make a
more specific hearsay objection, one the Michigan Court of Appeals found would have been
successful. But that Stefanski’s counsel admittedly made a mistake in failing to make a meritorious
argument for exclusion does not automatically satisfy the first prong of an ineffective assistance
of counsel claim. Strickland’s definition of deficient performance does not cover any mistake
made by an attorney; rather, it covers “errors so serious that counsel was not functioning as the
‘counsel’ guaranteed . . . by the Sixth Amendment.” Strickland, 466 U.S. at 687. Strickland warns
reviewing courts to avoid applying 20/20 hindsight “to conclude that a particular act or omission
of counsel was unreasonable.” Id. at 689. Considering counsel’s otherwise correct general hearsay
objection to the same testimony, “it is difficult to say that the admission of this one hearsay
statement demonstrates constitutionally deficient performance.” United States v. Robison,
205 F.3d 1342, 2000 WL 191852, at *8 (6th Cir. Feb. 11, 2000) (unpublished table decision); cf.
Meadows v. Doom, 450 F. App’x 518, 523 (6th Cir. 2011) (applying AEDPA and Strickland and
finding that, although “trial counsel could undeniably have been more vigilant in objecting” to
inadmissible statements, the errors were not so serious as to satisfy Strickland’s first prong). But
see Hurley, 426 F.3d at 376 (finding counsel’s “failure to object to any of the numerous improper
statements” made by the prosecutor outside the range of reasonableness).
Under the circumstances of this case, counsel’s single failure to cite the specific applicable
evidentiary rule—rather than the more general one—does not fall outside the “wide range of
professionally competent assistance.” Strickland, 466 U.S. at 690. This conclusion is supported
by our prior determination in Cathron v. Jones, 77 F. App’x 835 (6th Cir. 2003), where an attorney
- 18 - No. 23-1881, Stefanski v. Douglas
lodged an objection against the introduction of testimony on one ground but failed to object on a
stronger evidentiary ground. Id. at 843. Although the attorney should have lodged a more specific
objection, we did “not find that counsel’s failure to object was so unreasonable that he was not
functioning as counsel guaranteed by the Sixth Amendment.” Id. That reasoning is persuasive
here, and Stefanski does not cite case law counseling a different outcome. Accordingly, Stefanski
fails to establish deficient performance.
Assuming Stefanski could show deficient performance, he nonetheless fails to demonstrate
prejudice. At the outset, he is incorrect that the state court unreasonably applied federal law in
concluding he failed to prove prejudice. At the very least, reasonable jurists could disagree as to
whether the introduction of Lounds’s mother’s testimony prejudiced Stefanski. Richter, 562 U.S.
at 101. This renders relief inappropriate.
Stefanski argues that Lounds’s mother’s testimony was crucial to undercutting Stefanski’s
theory that Lounds fabricated the assaults to get a lenient sentence on his probation violation
because Lounds allegedly disclosed the assaults to his mother after he was already sentenced and
thus when he no longer had an incentive to report the assaults. In light of this timing, the state
court’s determination that Lounds’s mother’s testimony actually bolstered the motive-to-falsify
theory appears incorrect—at the point Lounds spoke to his mother, he no longer had the alleged
motive to lie. But this incorrect determination is not so unreasonable that it renders AEDPA
deference inappropriate. See Kayer, 592 U.S. at 118; Wiggins, 539 U.S. at 520–21. The state court
was correct that Lounds’s mother’s testimony was vague and did not provide any detail into
Lounds’s account of the underlying assaults, so it paled in comparison to the detail provided by
other testimony and renders it unlikely that the mother’s testimony played a prominent role in the
- 19 - No. 23-1881, Stefanski v. Douglas
jury’s decision. See Spaulding, 704 F. App’x at 481 (distinguishing between key testimony and
that which, “though helpful, . . . was far from critical to the prosecution’s case”).
As noted in the previous prejudice analysis, the testimony presented in this case differed
wildly. The government produced witnesses with accounts vastly different from the accounts
provided by Stefanski and his witnesses. The jury’s decision thus came down to which group to
believe; such analysis centered on credibility determinations. Although the timing of Lounds’s
disclosures certainly played a role in the case, it was not the only dispute, and perhaps not even
the main dispute. That Lounds’s mother’s statement helped the prosecution’s case does not render
it prejudicial—at least, not so prejudicial that it supports “a reasonable probability that,” but for
the erroneous introduction of the testimony, “the result of the proceeding would have been
different.” Strickland, 466 U.S. at 694. In speculating that a jury may reasonably believe that
Lounds would not have disclosed the sexual assaults to his mother if they were not true, Stefanski
offers just that—speculation. See Kayer, 592 U.S. at 118. Stefanski fails to show a “substantial”
likelihood of a different result to satisfy Strickland and falls far short of establishing that the state
court’s determination “lies ‘beyond any possibility for fairminded disagreement’” to overcome
AEDPA. Id. (internal citations omitted).
VI.
For the foregoing reasons, we affirm the district court’s rejection of Stefanski’s petition.
- 20 -