Michigan’s own rules of evidence already treat a patrol officer’s narrative as too unreliable to skip cross-examination. Charging, bond, and plea decisions run on it anyway, long before any court ever gets the chance.
A police report is hearsay: an out-of-court statement offered to prove the truth of what it describes. Michigan Rule of Evidence 803(8) denies police narrative reports the automatic pass that other government records get when offered against a defendant, and Michigan Compiled Laws 766.11b carves the police investigative report out of the hearsay exception it grants to almost every other kind of law enforcement paperwork.
The rules already build in the assumption that an officer’s account is not self-proving. Almost everything that happens to a defendant before trial, charging, bond, and the plea deal that resolves the overwhelming majority of cases, happens anyway on the strength of that same account, untested.
Key Points
Quick FAQs
What “Hearsay” Actually Means
Michigan Rule of Evidence 801 defines hearsay as a statement the declarant does not make while testifying at the current trial or hearing, offered to prove the truth of what it asserts. MRE 802 makes hearsay inadmissible unless another rule says otherwise. That is the whole doctrine. It is not a technicality about paperwork. It is a rule about who gets to be believed without being tested.
A police report is built almost entirely out of exactly that kind of statement. It records what the officer says happened, and it often records what a witness, a victim, or a suspect told the officer happened. None of that account was made under oath, in front of the person it will be used against, subject to cross-examination, at the moment it matters. That is the condition the hearsay rule exists to catch.
The Rule Already Treats Police Accounts as Different
Government records generally get a hearsay exception. MRE 803(8) lets in records of a public office’s activities, or a matter observed while under a legal duty to report, without making anyone testify to authenticate them. The premise is that a routine government record is made by someone with no stake in a particular outcome. A county clerk logging a filing is not trying to make a case against anyone.
“the office’s activities; or a matter observed while under a legal duty to report, but not including: in a criminal case, a matter observed by law-enforcement personnel”Michigan Rule of Evidence 803(8), public records exception
That carve-out is the whole argument in one sentence. The rule’s drafters did not treat a patrol officer’s observations during an active investigation the same as a routine government record, because a patrol officer investigating a crime is not a neutral record-keeper. The officer occupies an adversarial position relative to the person the report is about. The rule accounts for that. It denies the shortcut specifically where the shortcut would matter most.
Michigan Went Further at the Preliminary Exam Stage
Preliminary examinations already run on relaxed evidentiary standards, built for speed rather than a full evidentiary hearing. Michigan Compiled Laws 766.11b lets a magistrate admit several categories of hearsay without the author testifying: drug field-test results, certified court and government records, ordinary-course-of-business records, and, notably, most reports prepared by a law enforcement officer or other public agency.
Then the statute writes out the one category that would matter most.
“Except for the police investigative report, a report prepared by a law enforcement officer or other public agency.”MCL 766.11b(1)(d)
Lab results, forensic findings, and other law-enforcement-adjacent paperwork get the hearsay pass. The officer’s own narrative account of the investigation does not. Even in the venue Michigan built specifically to move fast and screen cases on a lower standard, the legislature decided the investigative report needed a live witness. The statute also lets either side subpoena the author of any admitted hearsay report on a satisfactory showing that live testimony would matter to the probable cause determination, which is itself an acknowledgment that these documents are not self-proving.
Pulling the rule text, the statute, and the underlying research behind a piece like this one takes real hours, and public records don’t request themselves. Clutch Confidential members fund that work directly.
$10/month ?Where the Rule Applies, and Where It Never Gets the Chance
I want to be direct about what the timeline actually shows, because it gets glossed over in most conversations about police accountability. No rule of evidence governs a charging decision. No rule of evidence governs a bond amount. No rule of evidence governs what a prosecutor offers at a plea conference. Those decisions run on the report, full stop, because nothing requires them not to.
The officer’s narrative account is created. Nothing in the rules of evidence governs how it gets written, and nothing tests it at this stage. It becomes the baseline document everything downstream is measured against.
A prosecutor reads the report and decides what to charge. This is not an evidentiary proceeding. The rules of evidence, including the hearsay rule, simply do not apply to a charging decision.
A magistrate sets pretrial conditions, often based directly on the report’s account of the underlying conduct and perceived risk. Bond hearings are not governed by the hearsay rule.
MCL 766.11b admits most law enforcement paperwork as hearsay here, but specifically excludes the police investigative report. If the state wants that narrative in, the officer generally has to testify and can be cross-examined on it, though the standard of proof is only probable cause.
This is where the overwhelming majority of cases resolve. No evidentiary hearing occurs. Both sides negotiate based on their read of the report, and the hearsay rule is never invoked because there is no proceeding for it to apply to.
MRE 802 bars the report as free-standing hearsay, and MRE 803(8) blocks the public records shortcut. The officer must testify live and face cross-examination for the narrative to become evidence. This is also the stage almost no case reaches.
The Counterargument: Doesn’t the Officer Usually Testify Anyway?
In the cases that reach a courtroom, yes, officers generally take the stand and can be cross-examined. That is the system working as designed. The problem is the assumption embedded in that answer: that the courtroom is where the case gets decided. For the vast majority of Michigan defendants, it isn’t. The plea agreement is negotiated, and the case is resolved, while the report is still an untested account that both sides are treating as accurate enough to build an outcome on. Cross-examination is a real safeguard. It is also a safeguard that arrives, on average, after the decision it would have tested has already been made.
What Happens When the Report Is Wrong
The hearsay rule assumes an officer investigating a case is not a neutral narrator. That is not an accusation. It is a structural observation the rule itself makes, and the exoneration data bears it out at scale.
The National Registry of Exonerations has found that official misconduct contributed to 54 percent of documented exonerations nationally, with police misconduct specifically present in 35 percent and prosecutorial misconduct in 30 percent. In homicide cases, the misconduct rate climbs to 72 percent. The single most common form of misconduct, appearing in 44 percent of cases, was concealing evidence that would have helped the defense. And when misconduct was later confirmed as a cause of a wrongful conviction, some form of discipline followed in only 17 percent of those cases. For prosecutors specifically, that figure drops to 4 percent.
The rule that excludes police observations from the easy path into evidence was not written because someone assumed officers lie. It was written because someone assumed officers are people, occupying an adversarial role, capable of being wrong in ways that a routine record-keeper isn’t. The exoneration numbers do not contradict that premise. They confirm it.
Police Are People. The Rule Already Knew That.
Nothing here argues that most officers fabricate their reports, and the data does not support that claim. Most documented misconduct in the exoneration record is not a story of officers inventing evidence from nothing. It includes tunnel vision on an early suspect, incomplete investigation, mistaken identification procedures, and pressure to close a case, alongside the smaller share of cases involving deliberate concealment or fabrication. Incompetence, institutional pressure, and misconduct produce the same downstream result: an inaccurate account, treated as accurate, because no one tested it before it did its work.
That is the entire structural logic behind excluding police observations from the routine hearsay exceptions. The rule does not need to resolve intent to justify skepticism. It only needs to recognize that a person writing an account of a case they are actively working, with an outcome they are institutionally invested in, is not the same as a clerk stamping a form. Michigan’s evidence rules got that right in the one venue built to test it. The rest of the system, where nearly every case is actually decided, never asks the question.
How Hard the Report Gets Tested, By Stage
The rule is honest about the report’s limits exactly once, at the trial almost no one’s case reaches. Every earlier decision point treats the same document as settled fact.
What Fixing This Would Actually Require
Diagnosing the gap is not the same as demanding an officer’s word count for nothing. The structural fix does not require abolishing reliance on police reports. It requires closing the gap between what the trial rule already assumes and what happens at every stage before trial.
That means requiring documented corroboration, not just a supervisor’s signature, before a report alone can support a felony charge. It means giving defense counsel funded, timely access to the underlying materials behind a report, not just the summary, before a plea is negotiated rather than after. It means treating pretrial detention decisions with the same skepticism about unverified officer narrative that MRE 803(8) already applies at trial. And it means building the kind of documented, public accountability record for sustained findings of officer misconduct that Michigan already requires for judges through the Judicial Tenure Commission, so that a charging decision, a bond hearing, and a plea negotiation can weigh an officer’s track record the same way a courtroom eventually would, instead of only after the case is already over.
None of that treats officers as adversaries. It treats the report the way the rule that governs it already does: as a document written by a person, in an adversarial position, that has not yet been tested. The system agrees with that assessment once a case reaches trial. The record shows it stops agreeing everywhere else.
Sources
How to Cite This Article
Continue Your Investigation
If this reporting raised more questions, use the Clutch Justice ecosystem to keep going.
