The same defect that ended the Flint prosecutions was present in Todd Robinson’s case. He was indicted by a judge who could not indict him and denied the hearing that should have followed. In February 2026 the Michigan Supreme Court explained why none of that helps him.
People v Robinson held that Peeler announced a new rule of law, that a Peeler violation does not implicate subject-matter jurisdiction, and that Peeler does not apply retroactively on collateral review. Applying the Linkletter-Hampton factors, the court reasoned that the rule concerns the process of getting to trial rather than the fairness of the trial itself, and that judges and prosecutors had relied on the prior unchallenged assumption. Relief was denied.
Why does subject-matter jurisdiction matter so much here?
Because a judgment entered without subject-matter jurisdiction is generally void and can be attacked at any time, regardless of finality or procedural default. Robinson argued the defective indictment deprived the trial court of jurisdiction. Had that argument succeeded, retroactivity analysis would have been unnecessary.
Is it unusual for a court to agree a rule is new and still deny relief?
No. Whether a rule is new and whether it applies retroactively are separate questions. Establishing novelty is what makes retroactivity analysis necessary in the first place, and that analysis frequently comes out against application to final cases.
Did the court say Robinson received a fair trial?
The Jackson Circuit Court had found that even if Peeler applied, Robinson could not show prejudice because a jury convicted him after a fair trial. The Supreme Court affirmed the denial of relief on alternative grounds resting on jurisdiction and retroactivity.
Could the Legislature change this outcome?
Retroactive relief is a question courts decide under this framework, but legislatures can create statutory avenues for post-conviction review. The repeal bill discussed in Part VI is prospective and does not address people already convicted.
The Case Against My Own Framing
I want to start by making the strongest version of the argument against where this installment ends up, because I think the court’s reasoning is more defensible than the outcome feels.
Finality is not an administrative convenience. A criminal judgment that can be reopened indefinitely is a judgment that never actually resolves anything, and the people who bear that uncertainty include victims and witnesses as well as defendants. Every legal system has to decide when a case is over, and any line drawn will leave someone on the wrong side of it.
The court also had a genuine analytical problem. Peeler was a statutory holding about which official may issue a charging document and what hearing must follow. That is not a rule about whether the jury heard reliable evidence or whether the defendant had counsel. A framework that asks whether a new rule goes to the fairness of the trial will not find much purchase in a rule about the route to trial.
So when the court concluded that the Peeler rule’s purpose is unrelated to the fairness of a trial, it was not being evasive. It was describing what Peeler actually decided.
That is the strongest case, and it is a real one. What follows is what it costs.
The Case
Todd Robinson was indicted in 2012 by a circuit judge sitting as a one-person grand jury. He moved for a preliminary examination and was denied, which was the correct result under Glass and under Green. He was convicted of first-degree premeditated murder.
After Peeler, he filed a successive motion for relief from judgment in Jackson Circuit Court, arguing that Peeler applied retroactively and that his indictment, and therefore his convictions, were void for lack of subject-matter jurisdiction. The circuit court denied the motion, finding Peeler did not apply retroactively and that in any event he could not show prejudice because a jury had convicted him after a fair trial.
The Court of Appeals affirmed, holding that Peeler did not affect subject-matter jurisdiction, did not announce a new rule of law, and did not apply retroactively on collateral review.
The Supreme Court affirmed in part and vacated in part. It held that Peeler did establish a new rule of law by barring one-person grand jury indictments and requiring preliminary examinations, vacating the Court of Appeals holding to the contrary. It also held that Peeler violations do not implicate subject-matter jurisdiction and do not apply retroactively on collateral review, affirming the denial of relief on those alternative grounds.
Robinson won the argument that the Court of Appeals had gotten wrong. The Supreme Court agreed with him that Peeler announced a new rule of law and corrected the lower court on that point. It did not change his position at all, because novelty was the threshold for an analysis he then lost on two independent grounds.
The Lab holds free Clutch Justice resources on locating Michigan appellate dockets, briefs, and opinions, including the case document archives where filings in matters like this one are published.
Explore The LabThe Two Doors, Both Closed
Robinson needed to get through one of two doors. The court closed both.
If the defective indictment deprived the trial court of subject-matter jurisdiction, the judgment would be void and finality would not protect it. The court held a Peeler violation does not implicate subject-matter jurisdiction.
This route required no retroactivity analysis at all. Closing it made the second question necessary.
Under the Linkletter-Hampton factors the court found the rule’s purpose unrelated to trial fairness, and that prosecutors and judges had relied on the prior assumption. Retroactive application was denied.
The characterization of the rule adopted here traces directly to the narrow statutory ground chosen in Peeler.
The Court of Appeals had wrongly held Peeler announced no new rule. That was vacated. The correction changed nothing about the disposition.
A ruling in his favor on the only question he won, with no effect on the outcome.
The Circularity in the Reliance Factor
One element of the analysis deserves closer attention than it has received, and I want to be careful to state it as an observation about the framework rather than as a claim that the court misapplied it.
The second Linkletter-Hampton factor asks about general reliance on the old rule. The court found this weighed against retroactivity, noting that it is undisputed that before Peeler, prosecutors and judges relied upon the unchallenged assumption that the one-man grand jury law allowed judges to issue criminal indictments.
Set that beside what Peeler held. The old rule was not a rule. It was an assumption that had never been examined, which the Supreme Court described in exactly those terms and then found to be wrong.
So the reliance that protects the convictions is reliance on a proposition the court has determined was never law. The longer the assumption went unexamined, and the more convictions accumulated under it, the stronger the case for leaving those convictions undisturbed.
A practice that persisted because nobody challenged it becomes, by virtue of that persistence, a reason not to disturb what it produced. Duration converts into justification, and the conversion happens without anyone deciding it should.
That is not a criticism of the Robinson court, which applied a framework it did not invent and which serves purposes I described at the top of this piece. It is an observation about what the framework does when the old rule was never actually a rule.
The Number Nobody Has
The third factor concerns the effect on the administration of justice, and it is here that the record shows something worth pausing on.
The court acknowledged that the exact number of defendants subjected to the errors identified in Peeler is unknown, while observing that the practice appears to have been used with some regularity over many decades in some courts. The Attorney General’s office has described the mechanism as producing thousands of prosecutions over a century.
Nobody knows. Not the court, not the Attorney General, not the county prosecutors, not the Legislature considering repeal.
There is no field in any Michigan case management system recording that a prosecution began with a one-person grand jury indictment. No county maintains a list. No audit has been performed. A person serving a sentence today may or may not have been charged this way, and neither they nor their counsel has any systematic way to find out.
The affected population cannot be identified, which means it cannot be notified, which means even the limited relief theoretically available to someone still in a direct appeal posture reaches only those who happen to learn of it. The absence of a count is not merely a research problem. It is the reason the ruling’s practical scope is set by chance.
Why This Matters
Put the two decisions side by side and the distribution is complete.
Three former state officials were charged through a mechanism the Supreme Court held could not produce a valid indictment. Their cases were pending on interlocutory appeal when the court so held. The prosecutions ended.
Todd Robinson was charged through the same mechanism, denied the same hearing, and convicted. His case was final when the court so held. He remains convicted.
Nothing distinguishes the two indictments. Both were issued by judges without the authority to issue them. What distinguishes the outcomes is where each case sat on the calendar in June 2022, and that is not a fact about either case. It is a fact about timing.
Behind Robinson is an unknown number of people in the same position, and the reason the number is unknown is that Michigan never built the record that would have made counting possible. The mechanism operated for a century, in closed proceedings, leaving no trace in any system designed to be searched.
That is the second half of this series. Not what the courts decided, but what cannot be seen.
Part V takes up MCL 767.4, which makes disclosure of information presented to a one-person grand jury a misdemeanor. A century of proceedings, the Flint evidence among them, sits behind that provision. Michigan now has a category of final convictions produced by a mechanism the Supreme Court has held was never authorized, arising from proceedings that cannot lawfully be examined.
Continue Your Investigation
If this reporting raised more questions, use the Clutch Justice ecosystem to keep going.
