AEDPA Deference After Conclusory State-Court Merits Denials: Richter “Could-Have-Supported” Review and Plea Validity Based on Statutory Maximum Notice

1. Introduction

This habeas appeal arises from Eric Lamont Thomas’s 2014 no-contest plea in Michigan state court to first-degree criminal sexual conduct, stemming from a 1996 home invasion and rape later linked to him through DNA testing. Before trial, Thomas rejected a plea offer calling for an 8–25 year sentence. After his codefendant pleaded and was allowed to testify, Thomas entered a no-contest plea that carried a statutory maximum of life imprisonment.

In post-conviction proceedings under Michigan Court Rules subchapter 6.500, Thomas asserted (i) due process defects because he allegedly did not understand his sentencing exposure when pleading, and (ii) ineffective assistance of trial counsel for allegedly failing to explain (a) sentencing exposure and (b) the practical scope of his appellate-waiver consequences. Michigan courts denied relief, and federal courts (under AEDPA) denied habeas relief.

The Sixth Circuit’s opinion addresses three questions central to modern habeas review of guilty-plea challenges: (1) whether sparse state-court reasoning still counts as a merits adjudication triggering AEDPA deference; (2) when a federal court applies Wilson v. Sellers versus Harrington v. Richter; and (3) what constitutes “clearly established” Supreme Court law on plea knowledge of sentencing consequences and counsel’s duties during plea advice.

2. Summary of the Opinion

The court concluded:

• AEDPA deference applied because the state trial court identified and rejected the federal due process and ineffective-assistance claims, even though its reasoning was sparse.
• Richter governed the federal court’s analysis because the state trial court did not provide “specific reasons” for denial; therefore the federal habeas court could consider arguments that “could have supported” the state decision.
• Ineffective assistance—appeal-waiver advice: No clearly established Supreme Court precedent required counsel to advise on the “nuances” that a plea waives appellate review of non-jurisdictional defects even by leave; thus AEDPA barred relief.
• Ineffective assistance—sentencing exposure: The record permitted a fairminded conclusion that Thomas understood his exposure (at least the statutory maximum) and was not credibly surprised; AEDPA barred relief.
• Due process—sentencing exposure: The state court could reasonably treat knowledge of the statutory maximum as sufficient under Supreme Court precedent; AEDPA barred relief.

3. Analysis

3.1. Precedents Cited

A. AEDPA, merits adjudications, and “look-through” doctrine

• Fleming v. Metrish: Provided the appellate standards (de novo for legal conclusions, clear error for factfinding) in federal habeas appeals.
• Stermer v. Warren: Framed the gateway question—AEDPA deference applies only to claims “adjudicated on the merits.” Also supplied the “look-through” approach later operationalized via Wilson v. Sellers.
• Jones v. Shoop (quoting Moss v. Miniard): Reiterated that federal courts consult the last reasoned state-court opinion.
• Harrington v. Richter: Central here. It created a presumption that a state court adjudicated a federal claim on the merits when it denies relief, absent indications to the contrary; and it instructed federal courts to consider arguments that “could have supported” an unexplained merits denial.
• Johnson v. Williams: Explained limited circumstances for rebutting Richter’s merits presumption (e.g., a federal claim was inadvertently overlooked).
• Smith v. Cook: Applied Richter even when a state opinion omits discussion of a federal claim or addresses some but not all claims—supporting the Sixth Circuit’s conclusion that the Michigan trial court adjudicated the relevant claims on the merits.
• Wilson v. Sellers: Established that when the last state court provides a reasoned opinion, federal courts review and defer to the “specific reasons given.” The Sixth Circuit distinguished Wilson because the state trial court’s language (“no error attributed to trial counsel”) lacked “specific reasons.”

B. The “unreasonable application” bar and deference intensity

• Mays v. Hines (quoting Harrington v. Richter): Emphasized that “unreasonable” under AEDPA means more than error; it requires “extreme malfunctions.”
• Yarborough v. Alvarado (quoted in Harrington v. Richter): “Fairminded jurists could disagree” is enough to deny habeas relief.
• Cullen v. Pinholster (quoting Woodford v. Visciotti): Reiterated the “highly deferential” posture AEDPA demands.
• Parker v. Matthews: Reaffirmed that circuit precedent (even Sixth Circuit precedent) is not “clearly established Federal law, as determined by the Supreme Court.”

C. Plea-stage ineffective assistance (Strickland/Hill) and “collateral consequences”

• Strickland v. Washington: Governing standard—deficient performance plus prejudice.
• Hill v. Lockhart: Extended Strickland to guilty pleas, but (as later noted) did not resolve counsel’s duties as to “collateral consequences.”
• Knowles v. Mirzayance and Harrington v. Richter: Produced the “doubly” deferential review on AEDPA plus Strickland.
• Chaidez v. United States: Noted that Hill left open whether advice about collateral consequences must satisfy the Sixth Amendment; it also reflected that many courts historically held there is no constitutional duty to advise of collateral consequences.
• Padilla v. Kentucky: Required advice about deportation risk but characterized deportation as “unique,” and declined to use the direct/collateral distinction to define the entire scope of Strickland. The Sixth Circuit leveraged Chaidez and Padilla to support that (under AEDPA) it would not be “clearly established” that counsel must advise on the nuanced scope of appellate waiver consequences.
• United States v. Morris: Invoked by Thomas to argue for broader record review; the Sixth Circuit rejected it as non-Supreme-Court authority and thus not a basis for AEDPA relief.

D. “Clearly established” law and general principles

• Andrew v. White: Held that general legal principles can qualify as “clearly established” law under AEDPA when sufficiently fundamental. Thomas attempted to derive such a principle from Strickland; the court rejected the attempt because it pointed to a generic fragment of Strickland’s methodology rather than a fundamental rule compelling a specific outcome.
• Payne v. Tennessee and White v. Woodall: Used in Andrew’s explanation of when general principles are “fundamental enough” to dictate results across new factual permutations.

E. Due process standards for plea validity and sentencing awareness

• Boykin v. Alabama: Required an affirmative showing that a plea is intelligent and voluntary; footnote guidance suggested courts are “best advised” to confirm understanding of “the permissible range of sentences,” quoting Commonwealth ex rel. West v. Rundle.
• Brady v. United States: Required that a plea be voluntary, knowing, and intelligent with “sufficient awareness of the relevant circumstances and likely consequences.”
• Ruelas v. Wolfenbarger: Although not “clearly established” Supreme Court law, it illustrated a reasonable interpretation of Boykin/Brady—that awareness of the statutory maximum can satisfy due process for plea consequences.
• Carey v. Musladin: Reinforced that where Supreme Court precedent does not clearly dictate the claimed rule, state courts retain leeway; the Sixth Circuit used this to explain why a state court could reasonably interpret “likely consequences” to require, at most, knowledge of the statutory maximum.

F. Michigan sentencing context (background) and state post-conviction pathway

• King v. Berghuis, People v. Drohan (overruled in part on other grounds by Alleyne v. United States), and People v. Babcock: Explained Michigan’s indeterminate sentencing scheme and the (then) mandatory nature of Michigan sentencing guidelines.
• People v. McSwain: Described MCR 6.500 as the exclusive post-appeal vehicle to challenge convictions in Michigan.
• People v. Thomas (Michigan Supreme Court mem. orders): Demonstrated the state appellate disposition pattern—summary denials that required federal “look-through” to the trial court, and then (because of minimal reasoning) resort to Richter’s “could-have-supported” analysis.
• Thomas v. Burt (district court and prior Sixth Circuit decisions referenced as Thomas I and Thomas II): Provided procedural history and clarified which claims survived to this appeal.

3.2. Legal Reasoning

A. Why AEDPA deference applied

The Sixth Circuit treated the Michigan trial court’s denial as a merits adjudication because the court expressly recognized Thomas’s claims as federal constitutional due process and ineffective-assistance contentions and rejected them (“no error attributed to trial counsel” and “no defect in the plea proceedings”). Under Harrington v. Richter and Smith v. Cook, even sparse or incomplete reasoning does not defeat the presumption of a merits adjudication.

B. Why the case turned on Richter rather than Wilson

The doctrinal fulcrum is the distinction between:

• Wilson v. Sellers: Defer to “specific reasons given” in a reasoned state-court opinion.
• Harrington v. Richter: If the state decision is unexplained (or effectively unexplained), federal courts ask what rationales “could have supported” the outcome, and deny relief if any such rationale is consistent with Supreme Court holdings.

Although the state trial court issued multiple pages, the Sixth Circuit characterized its treatment of the key federal issues as effectively conclusory—too perfunctory to qualify as “specific reasons” within Wilson’s framework. That classification expanded the range of permissible state-court rationales under AEDPA and made Thomas’s path to relief significantly steeper.

C. Ineffective assistance—appeal-waiver advice (claim 6)

Thomas alleged counsel failed to advise that the plea would waive appellate review of non-jurisdictional issues even by leave to appeal. The Sixth Circuit did not decide the best view of professional norms on de novo review; instead it asked whether Supreme Court precedent clearly established such a duty.

Relying on Hill v. Lockhart, Chaidez v. United States, and Padilla v. Kentucky, the court emphasized that the Supreme Court has not clearly established a general constitutional obligation to advise about all collateral consequences of a plea, and that Padilla carved out deportation due to its “unique nature.” Under AEDPA, that gap meant a “fairminded jurist” could deny the claim without unreasonably applying Supreme Court law.

The court also rejected Thomas’s attempt to use Andrew v. White to convert general Strickland language into a clearly established rule compelling relief, holding that the cited snippet was not a “fundamental” principle that made the demanded application “beyond doubt.”

D. Ineffective assistance—sentencing exposure (claim 5)

The court accepted as a permissible rationale that the record supports a finding Thomas understood his exposure: the Plea Form listed “Life” as the statutory maximum; the plea colloquy confirmed he understood the maximum; and at sentencing he confirmed reviewing the presentence report and guidelines with counsel and raised no objection to guideline calculations. In addition, his later affidavit focused on appellate issues rather than surprise about sentencing exposure, which could be viewed as undermining credibility on this point.

Under the “doubly” deferential Strickland + AEDPA framework (Harrington v. Richter; Knowles v. Mirzayance), these record-based inferences were enough to defeat habeas relief.

E. Due process—sentencing exposure (claim 2)

Thomas urged that due process requires understanding of guideline exposure (not merely the statutory maximum), invoking Boykin v. Alabama and Brady v. United States. The Sixth Circuit held that Supreme Court law does not clearly require the more detailed understanding Thomas sought; rather, it was reasonable to interpret Brady’s “likely consequences” and Boykin’s “permissible range” (in the relevant footnote guidance) as satisfied by knowledge of the statutory maximum.

The court buttressed “reasonableness” by referencing Ruelas v. Wolfenbarger as an example of a plausible (albeit non-controlling) interpretation consistent with that view, and invoked Carey v. Musladin to underscore that the absence of a directly controlling Supreme Court rule defeats a claim under AEDPA’s “clearly established” requirement.

3.3. Impact

A. Practical habeas impact: classifying state opinions as “unexplained”

The opinion meaningfully reinforces that even multi-page state post-conviction orders may be treated as effectively “unexplained” for AEDPA purposes if they do not supply “specific reasons” on the federal claim. That classification triggers Richter’s “could-have-supported” approach, which is often decisive because it allows federal courts to deny relief so long as any reasonable theory consistent with Supreme Court precedent can justify the result.

B. Plea litigation: statutory maximum notice as a safe harbor under AEDPA

For due process challenges premised on misunderstanding sentencing exposure, the decision signals that—at least on federal habeas review—state courts have wide latitude to deem plea validity satisfied when the defendant is advised of, and acknowledges understanding, the statutory maximum penalty, even if guideline mechanics or likely minimums are not fully explored on the record.

C. Ineffective assistance: appellate-waiver “nuances” remain hard to constitutionalize on habeas

The decision also illustrates the difficulty of converting best practices (advising on practical appeal forfeitures) into AEDPA-enforceable constitutional rules absent a directly controlling Supreme Court decision. Even where the record suggests the defendant misunderstood the scope of appellate waiver—as the court itself noted in prior proceedings (Thomas II)—AEDPA may foreclose relief if Supreme Court precedent has not clearly established the specific obligation asserted.

D. Incentive effects in state practice

Although the ruling is formally about federal habeas limits, it implicitly incentivizes state trial courts and defense counsel to build robust plea records (clear explanation of appellate consequences and sentencing exposure) to reduce post-conviction litigation, even if AEDPA will later shield many state outcomes.

4. Complex Concepts Simplified

• AEDPA deference (28 U.S.C. § 2254(d)): Federal courts generally cannot grant habeas relief just because they think the state court was wrong; the state decision must be unreasonably wrong under Supreme Court precedent (not circuit law).
• “Adjudicated on the merits” presumption (Harrington v. Richter): If a state court denies relief after the federal claim is presented, federal courts presume it decided the claim on its substance unless clearly shown otherwise.
• Wilson v. Sellers vs. Harrington v. Richter: If the last reasoned state decision provides real reasons, federal courts evaluate those reasons (Wilson). If the state decision gives no meaningful reasoning, federal courts ask what arguments could have supported the result (Richter).
• “Clearly established Federal law”: Only Supreme Court holdings count. Lower-court decisions (including Sixth Circuit cases) can be persuasive but cannot by themselves make a rule “clearly established.”
• Direct vs. collateral consequences: Some plea consequences (like the immediate sentencing range) are “direct.” Others (like many post-plea effects) have been treated as “collateral.” The Supreme Court has not clearly imposed a broad duty to advise on all collateral consequences, with Padilla v. Kentucky treating deportation as “unique.”
• Michigan indeterminate sentencing: Defendants receive a minimum and maximum term; guidelines largely set the minimum, while statutes often set the maximum. Parole eligibility begins after the minimum, but release before the maximum is discretionary.

5. Conclusion

The Sixth Circuit’s opinion underscores a core reality of modern federal habeas review: the decisive battleground is often not whether a plea was ideally explained, but whether Supreme Court precedent clearly compels the petitioner’s rule and whether any reasonable rationale can support the state court’s denial under AEDPA. By treating the state post-conviction order as effectively unexplained, the court applied Harrington v. Richter’s “could-have-supported” framework, rejecting ineffective-assistance claims about appellate-waiver nuances and sentencing advice, and rejecting a due process claim seeking more than notice of the statutory maximum. The case thus strengthens the protective moat AEDPA builds around state guilty-plea convictions—particularly when state courts issue terse denials and the Supreme Court has not squarely addressed the asserted constitutional duty.