Nissenbaum v. Jennings (7th Cir. 2025):
Clarifying “Reasonable Probability” for Lost-Plea Ineffective-Assistance Claims under AEDPA

1. Introduction

The Seventh Circuit’s decision in Nathan Nissenbaum v. Chad Jennings addresses the recurring tension between state-court findings, federal habeas deference under the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), and the Sixth Amendment right to effective counsel during plea bargaining. The petitioner, Nathan Nissenbaum, contended that his trial lawyer’s deficient advice caused him to forgo a favourable plea and ultimately receive a 14-year sentence after conviction. At stake were two key issues:

• Whether the Illinois Appellate Court unreasonably applied Strickland v. Washington, Missouri v. Frye, and Lafler v. Cooper when it found no prejudice.
• How federal courts should conduct AEDPA review when the final reasoned state-court opinion decides only one Strickland prong (prejudice) and leaves the performance prong untouched.

Although the Seventh Circuit ultimately denied relief, its opinion furnishes important guidance on (a) what “reasonable probability” means when the plea offer never crystalised into a formal, signed proposal, and (b) how to navigate AEDPA’s deferred-prong puzzle while remaining within the Supreme Court’s instruction in Wilson v. Sellers.

2. Summary of the Judgment

Writing for a unanimous panel, Judge Kirsch affirmed the district court’s denial of habeas corpus. Key points:

• AEDPA Deference: The federal court reviews only the state appellate court’s prejudice determination because that was the “last reasoned decision.”
• No Unreasonable Application: The state court’s conclusion—that Nissenbaum failed to show a reasonable probability he would have accepted, and could have finalised, a plea before it was withdrawn—was not so wrong that “no fair-minded jurist could agree.”
• Performance Prong Not Reached: Because the prejudice prong failed under AEDPA, the panel declined to decide whether counsel’s performance was deficient, leaving the Thomas/Wilson tension for another day.

3. Detailed Analysis

A. Precedents Cited

1. Strickland v. Washington, 466 U.S. 668 (1984) – established the two-prong (performance & prejudice) test for ineffective assistance.
2. Missouri v. Frye, 566 U.S. 134 (2012) – clarified prejudice where counsel allowed a plea offer to lapse.
3. Lafler v. Cooper, 566 U.S. 156 (2012) – recognised prejudice where bad advice prompted rejection of a plea actually offered.
4. Wilson v. Sellers, 584 U.S. 122 (2018) – adopted a rebuttable presumption that higher state courts silently adopt lower-court reasoning.
5. Harrington v. Richter, 562 U.S. 86 (2011) & Wiggins v. Smith, 539 U.S. 510 (2003) – set AEDPA’s “fair-minded disagreement” ceiling.
6. Seventh Circuit internal precedent – Thomas v. Clements, 789 F.3d 760 (7th Cir. 2015) & Woolley v. Rednour, 702 F.3d 411 (7th Cir. 2012) about selecting the “last reasoned decision.”

The Seventh Circuit acknowledged but sidestepped a potential conflict between Thomas and Wilson, signalling that future panels may revisit the “look-through” question.

B. The Court’s Legal Reasoning

1. Step 1 – Identify the State Decision to Review
   • The Illinois trial court decided both Strickland prongs; the appellate court affirmed only on prejudice.
   • Following Seventh Circuit precedent (Thomas), AEDPA deference attaches solely to the prejudice analysis in that last reasoned opinion.
2. Step 2 – Check for Legal Error
   • Petitioner argued the state court conflated “reasonable probability” with a heightened “manifest-weight-of-the-evidence” standard.
   • The panel found only loose phrasing, not an actual misapplication of law; the court had recited Frye/Lafler correctly.
3. Step 3 – Assess Factual/Legal Reasonableness
   • No Formal Offer. The prosecutor’s March 24 text (“could probably make an offer”) was not a formal plea. Still, the state court assumed arguendo that a deal was possible.
   • Timeline Problem. The opening evaporated 18 days later when the prosecutor consulted the victim. The court held it speculative that negotiations could have concluded and the judge accepted a plea within that narrow window.
   • Would Petitioner Have Accepted? Contemporaneous evidence (emails, meetings, defendant’s dislike of any felony label) showed reluctance, undermining a “reasonable probability” of acceptance.
   • Trial-Risk Assessment. Because the victim’s cooperation was uncertain, an acquittal remained plausible; thus proceeding to trial was not irrational.
4. Step 4 – AEDPA Lens
   Even if another judge might credit the defendant’s hindsight, AEDPA asks whether all fair-minded jurists must reject the state court’s view. That high threshold was not met.

C. Impact of the Decision

• Clarifies “Reasonable Probability” Where No Formal Plea Existed
  • Defendants must do more than show interest; they must demonstrate a realistic path to consummating an agreement before it disappears.
• AEDPA Deference Reinforced
  • Reiterates that federal courts cannot second-guess state credibility determinations absent “clear and convincing” rebuttal.
  • Escalates the evidentiary burden for prisoners alleging lost-plea prejudice.
• Pending Tension on the “Look-Through” Doctrine
  • The panel expressly leaves unresolved whether Thomas survives Wilson when the higher state court reasons only on one prong of a mixed question.
  • Future litigants should brief this structural issue; Supreme Court guidance may follow.
• Plea-Bargaining Practice
  • Defence counsel should create written memorials of clients’ sentencing-range understanding and plea decisions.
  • Prosecutors may note that tentative conversations—until formalised—give little habeas exposure.

4. Complex Concepts Simplified

AEDPA Deference (§2254(d))

The federal court can overturn a state decision only if it is so wrong that reasonable judges could not agree with it (“unreasonable,” not merely “incorrect”).

Reasonable Probability

More than a mere possibility but less than “more likely than not.” It must undermine confidence in the outcome.

Manifest Weight of the Evidence (Illinois)

A deferential state-appellate review asking whether the trial court’s factual finding is “palpably erroneous.” Federal courts do not adopt this standard but recognise its use by state courts.

Look-Through Presumption (Wilson v. Sellers)

When a higher state court issues a summary ruling, federal courts “look through” to the last reasoned decision to see the rationale they must evaluate under AEDPA.

5. Conclusion

Nissenbaum v. Jennings does not blaze an entirely new doctrinal trail, but it sharpens the contours of existing precedent in two respects. First, it shows how difficult it is, under AEDPA, to convert alleged plea-bargaining misadvice into federal habeas relief when the “offer” was embryonic and the defendant’s documented behaviour contradicts his after-the-fact narrative. Second, the case highlights an unresolved methodological split on whether federal habeas courts must “look through” when the last state opinion only partially addresses Strickland. Practitioners should heed the opinion’s lessons: preserve written advisals, pursue concrete plea terms quickly, and recognise that afterthought assertions will rarely surmount AEDPA’s high bar. In the broader context, the decision reinforces the judiciary’s insistence that effective assistance claims in the plea arena be grounded in contemporaneous facts, not retrospective speculation.