Michigan Rejects Oregon v Kennedy Under Const 1963, art 1, § 15 and Adopts Pool v Superior Court to Bar Retrial After Prosecutorial-Misconduct Mistrials

Case: People of Michigan v. Devante Kyran Jennings (Mich Sup Ct, July 24, 2026)

1. Introduction

People v Jennings arises from a vehicle stop following a shooting investigation in Clinton Township. Police found a loaded handgun in the glovebox. During a custodial interview, Devante Kyran Jennings admitted he drove the vehicle and had been at the scene, but he refused a requested DNA swab and ended the interview—invoking his right to remain silent.

At Jennings’s first trial for carrying a concealed weapon, MCL 750.227, the prosecutor argued in closing that Jennings’s decision to end the interview showed a “guilty conscience” and contrasted Jennings with other occupants who continued speaking with police. The trial judge concluded the prosecutor had “weaponized” Jennings’s invocation of silence, granted a defense-requested mistrial, and later allowed retrial under the federal standard of Oregon v Kennedy, 456 US 667 (1982) (requiring intent to “goad” a mistrial).

After Jennings was convicted at a second trial, the key issue became constitutional and systemic: Does Const 1963, art 1, § 15 bar retrial more broadly than Kennedy when a mistrial is granted due to prosecutorial misconduct?

2. Summary of the Opinion

Under the adopted Pool framework, retrial is barred when: (1) a mistrial is granted because of improper prosecutorial conduct; (2) the conduct, taken as a whole, constitutes intentional conduct the prosecutor knows is improper and prejudicial, pursued for an improper purpose with indifference to a significant danger of mistrial or reversal (not merely negligence, mistake, or insignificant impropriety); and (3) the prejudice cannot be cured by means short of mistrial.

The Court vacated the Court of Appeals judgment and remanded for the trial court to apply the new state constitutional standard, emphasizing that the record was not developed for Pool findings.

3. Analysis

3.1 Precedents Cited

A. Federal double-jeopardy foundations and mistrial doctrine

• Green v United States, 355 US 184 (1957): supplied the Court’s lodestar articulation of the anti-harassment and anti-repeated-prosecution purposes of double jeopardy (“embarrassment, expense and ordeal,” “anxiety and insecurity”). The majority used Green to argue that a goading-only rule fails to protect the same interests when the prosecutor’s goal is to prejudice, avoid acquittal, or harass.
• Wade v Hunter, 336 US 684 (1949) and United States v Jorn, 400 US 470 (1971): supported the “valued right to have his trial completed by a particular tribunal.” Jorn also framed the concept of “prosecutorial or judicial overreaching” as a limit on retrial even when the defendant moves for mistrial.
• United States v Perez, 22 US (9 Wheat) 579 (1824) and Arizona v Washington, 434 US 497 (1978): anchored the “manifest necessity” doctrine (mistrial over defense objection) and the prosecution’s “heavy burden” to justify retrial in that setting—background used to explain why the system is especially vulnerable when prosecutors can effectively force defense mistrials without meaningful double-jeopardy consequence.
• United States v Tateo, 377 US 463 (1964) and Burks v United States, 437 US 1 (1978): captured the balancing principle (society’s interest in punishing guilt vs. the general bar on retrials) and the anti-“second chance to supply evidence” rationale.

B. The pre-Kennedy “bad faith/overreaching” line and Michigan’s pre-Kennedy posture

• United States v Dinitz, 424 US 600 (1976) and Lee v United States, 432 US 23 (1977): described a broader protection than goading alone—barring retrial where misconduct was intended to provoke mistrial or to “harass or prejudice” the accused. The majority relied on this history to argue that the ratifiers of the 1963 Constitution would have understood double jeopardy to guard against intentional manipulations beyond goading.
• People v Anderson, 409 Mich 474 (1980) and People v Benton, 402 Mich 47 (1977): showed Michigan’s pre-Kennedy embrace of the “bad faith” framing, supporting the majority’s claim that Michigan’s own jurisprudential trajectory was broader than the later federal contraction.

C. Oregon v Kennedy (the rejected federal standard)

• Oregon v Kennedy, 456 US 667 (1982): held retrial is barred only when prosecutorial conduct was intended to provoke the defense into moving for a mistrial. The Michigan Supreme Court rejected Kennedy under state law as “unduly rigid,” insufficiently protective of Article 1, § 15’s purposes, and practically “illusory” because proving subjective intent to goad is “virtually impossible” without a concession.
• The majority leaned heavily on the separate writing in Kennedy by Justice Stevens (concurring in the judgment), quoted and discussed as an internal critique of the goading-only rule and as conceptual support for a broader, purpose-driven approach.

D. State constitutional departures and the adopted standard

• Pool v Superior Court, 139 Ariz 98 (1984): became the controlling Michigan constitutional test. The majority viewed Pool as both (i) more consistent with double-jeopardy purposes (anti-harassment, tribunal-finality, anti-gamesmanship) and (ii) more administrable than the subjective goading inquiry because it permits objective inference from circumstances and treats prosecutorial knowledge of impropriety as presumable at some point.
• The opinion also surveyed other states as confirmatory reasoning (not binding): State v Kennedy, 295 Or 260 (1983), State v Breit, 122 NM 655 (1996), Thomas v Eighth Judicial Dist Court in & for Co of Clark, 133 Nev 468 (2017), Rogan, 91 Hawai'i 405 (1999), Batts, 30 Cal 4th 660 (2003), and Pennsylvania authority including Commonwealth v Johnson, 659 Pa 277 (2020).

E. Michigan constitutional interpretation and “lockstepping” cases

• People v Nutt, 469 Mich 565 (2004), People v Davis, 472 Mich 156 (2005), and People v Smith, 478 Mich 292 (2007): were central to the majority’s interpretive method. The majority accepted these cases for a limited proposition: the ratifiers intended to incorporate then-existing federal principles at ratification, but not to pre-approve all future federal doctrinal changes.
• People v Bullock, 440 Mich 15 (1992), Sitz v Dep't of State Police, 443 Mich 744 (1993), and People v Tanner, 496 Mich 199 (2014): framed the debate over whether a “compelling reason” is necessary to depart from federal analogues. The majority treated “compelling reason” as not a thumb on the federal scale and emphasized independent state constitutional interpretation keyed to ratifier intent at the time of adoption.
• The dissent (Zahra, J., joined by Bernstein, J.) argued the majority improperly cast aside People v Dawson, 431 Mich 234 (1988) and Michigan’s lockstep tradition, and faulted the majority for not conducting a stare decisis analysis.

3.2 Legal Reasoning

A. The majority’s core move: temporal ratifier-intent “anchoring”

The majority’s reasoning hinges on a temporally bounded version of lockstepping: because the ratifiers of Const 1963, art 1, § 15 looked to then-existing federal principles, federal doctrines that were “clear” at or before 1963 may be presumed incorporated; doctrines “first announced after ratification” carry no such presumption. Since Oregon v Kennedy (1982) post-dates 1963 by roughly two decades and significantly narrowed the preexisting “bad faith/overreaching” discourse, the Court held that mechanically adopting Kennedy would improperly delegate Michigan constitutional meaning to later federal reinterpretations.

B. Purpose-based critique of Kennedy

The majority treated double jeopardy as protecting substantive interests (freedom from repeated prosecutions and harassment; completion of trial before the first tribunal; prevention of state gamesmanship). On that view, a goading-only rule underprotects defendants where prosecutors commit intentional, prejudicial misconduct to avoid acquittal or to gain tactical advantage—because the defendant’s “choice” between a tainted trial and a mistrial is equally “hollow” even without goading.

C. Why Pool fits the majority’s conception of Article 1, § 15

The adopted Pool test is presented as a calibrated middle ground:

• Not strict liability: it does not bar retrial for “legal error, negligence, mistake, or insignificant impropriety.”
• Intent-focused: it targets intentional conduct known to be improper and prejudicial, pursued for an improper purpose, coupled with indifference to a significant risk of mistrial/reversal.
• Remedy-sensitive: it requires incurable prejudice—prejudice that “cannot be cured by means short of a mistrial.”
• Objective inference: intent/knowledge may be inferred from “objective factors,” including trial circumstances and the strength of the case/danger of acquittal.

The Court also emphasized that defendants still carry a “heavy burden” and that its holding should not be read to constitutionalize every prosecutorial misstep.

D. Remand and record-development

Because the trial court applied Kennedy, it did not make the specific findings demanded by Pool (knowledge of impropriety, improper purpose, indifference to mistrial/reversal risk, curability). The Supreme Court therefore vacated and remanded for trial-court application of the new state constitutional standard, permitting briefing and an evidentiary hearing, while instructing the court to rely “primarily” on objective circumstances.

E. The dissent’s competing theory

Justice ZAHRA’s dissent advanced a lockstep-plus-stare decisis position: that People v Dawson, 431 Mich 234 (1988) adopted Kennedy for Michigan; that post-1963 cases (including People v Nutt, 469 Mich 565 (2004)) show ratifiers intended coterminous interpretation; and that the majority both lacked a “compelling reason” and failed to justify overruling precedent. The dissent further argued this case was a poor vehicle because, in its view, neither Kennedy nor Pool would be satisfied on the record.

3.3 Impact

A. Immediate doctrinal change in Michigan

Michigan now recognizes, under Const 1963, art 1, § 15, a broader double-jeopardy bar to retrial after defense-requested mistrials than the federal Fifth Amendment minimum established by Oregon v Kennedy. Trial and appellate courts must apply Pool rather than asking solely whether the prosecutor intended to “goad” the defense into a mistrial request.

B. Practical effects on trial practice and prosecutorial incentives

• Greater consequence for intentional “win-at-all-costs” misconduct: Prosecutors face a heightened constitutional risk if they knowingly deploy improper, highly prejudicial tactics in close cases (especially those that appear to be going badly) with indifference to mistrial/reversal.
• Record-making becomes critical: Courts should expect litigants to develop evidence and arguments on (i) what the prosecutor knew, (ii) whether the misconduct served an improper purpose, (iii) whether alternatives to mistrial could cure the prejudice, and (iv) objective contextual factors (including strength of evidence and risk of acquittal).
• MCR 6.417 and cure analysis gain constitutional importance: Because Pool requires incurable prejudice, courts will more explicitly analyze curative instructions, adjournments, and other alternatives before granting mistrial, and will make findings that later control whether retrial is barred.

C. Relationship to federal law

The decision does not alter federal double-jeopardy doctrine; it establishes a state constitutional floor above the federal minimum. Michigan courts will thus be applying a Michigan-specific standard in this mistrial context even when the federal claim fails under Kennedy.

D. Open questions the Court flagged

The majority confined its holding to the mistrial context and expressly left open whether Const 1963, art 1, § 15 may bar retrial when a conviction is reversed on direct appeal due to egregious prosecutorial misconduct.

4. Complex Concepts Simplified

• Double jeopardy: A constitutional protection against being prosecuted (or punished) more than once for the same offense. Here, the relevant strand is protection against successive prosecutions.
• When “jeopardy attaches”: Typically when the jury is selected and sworn (or, in a bench trial, when the court begins hearing evidence). Once it attaches, ending a trial without a verdict can trigger double-jeopardy limits on retrial.
• Mistrial: The trial ends without a verdict due to a serious error. A new trial may or may not be allowed depending on why the mistrial happened and who sought it.
• “Manifest necessity”: The high standard that can permit retrial when a mistrial is declared over the defendant’s objection (e.g., a deadlocked jury).
• Oregon v Kennedy “goading”: The federal rule that bars retrial only if the prosecutor intentionally tried to provoke the defendant into asking for a mistrial.
• Pool v Superior Court standard (now Michigan law): A broader rule that bars retrial when the mistrial was caused by prosecutorial misconduct that is intentional, known to be improper and prejudicial, pursued for an improper purpose with indifference to a significant risk of mistrial/reversal, and incurably prejudicial.
• “Objective factors”: Because a prosecutor’s internal thoughts are rarely provable, courts infer knowledge/intent from circumstances—what happened, how obvious the impropriety was, whether the case was going badly, whether the conduct was repeated, and whether the prosecutor ignored obvious limits.

5. Conclusion

People v Jennings is a major state-constitutional turning point in Michigan criminal procedure. The Court rejected the federal Oregon v Kennedy goading-only rule as inconsistent with Const 1963, art 1, § 15’s ratifier-intended protections against harassment, repeated prosecutions, and loss of the first tribunal. In its place, Michigan adopted Pool v Superior Court, making retrial constitutionally unavailable when a prosecutor’s intentional, knowingly improper and prejudicial misconduct—pursued for an improper purpose with indifference to mistrial/reversal risk—forces a mistrial that cannot be cured by lesser means.

The decision both increases the constitutional stakes of certain forms of intentional prosecutorial misconduct and demands more explicit trial-court findings on intent, knowledge, purpose, risk, and cure—issues that will now shape the permissibility of retrial in Michigan.