Michigan Judges Keep Citing ‘Deterrence.’ The State Has No Working Way to Check If It’s True.

Every Michigan sentence is legally required to weigh deterrence. Nothing in the state is legally required to check whether it worked, and right now, people are dying inside while that gap stays open.

The Doctrine Every Michigan Judge Has to Recite

I want to be direct about this because it gets treated as boilerplate: when a Michigan judge invokes deterrence at sentencing, that is not a rhetorical flourish. It is the law.

In 1972, the Michigan Supreme Court decided People v Snow, which set out four considerations every sentencing judge in the state must weigh: reformation of the offender, protection of society, disciplining of the wrongdoer, and, in the court’s language, “the deterrence of others from committing like offenses.” Michigan Court of Appeals opinions cite Snow constantly, in cases spanning armed robbery, arson, and juvenile sentencing. It is one of the most repeated doctrinal phrases in Michigan criminal law.

The doctrine treats deterrence as settled. The criminology does not.

What the Research Actually Says About Deterrence

The National Institute of Justice distilled decades of research from criminologist Daniel Nagin into a short public guidance document. Its core finding: “certainty of being caught is a vastly more powerful deterrent than punishment.” Nagin’s broader body of work goes further. Increasing the severity of punishment through longer sentences does little to deter crime on its own. People considering a crime respond far more to their perceived odds of getting caught than to how many years they might serve if they do. Short to moderate sentences may carry some deterrent value; longer sentences add very little on top of that.

That is a direct challenge to the fourth Snow factor as most judges apply it. Certainty of apprehension is a function of policing, investigation, and clearance rates. Severity of sentence is the one lever a judge controls at sentencing, and it is the lever the research says matters least.

Michigan’s Own Numbers Undercut the Theory

If severity of punishment reliably deterred crime, Michigan’s prison population and its crime trends should move together in a predictable way as sentencing got tougher or lighter. They do not.

From 1984 to 1989, Michigan’s prison population nearly doubled, from 14,658 to 31,834, following the repeal of an emergency overcrowding release law and a surge in drug-related commitments. In 1998 and 2000, the state’s Truth in Sentencing laws eliminated parole eligibility before a full minimum sentence and ended good-time credits, pushing the population to a record high of 51,554 in March 2007. Sentences got longer and more certain to be fully served throughout that entire climb.

Then the population fell 36 percent by 2024, down to 32,778. According to Michigan’s own corrections department, that decline tracked a 45.7 percent drop in statewide arrests between 2007 and 2023, plus policy changes to parole and probation violations, not a wave of newly deterred would-be offenders. The department’s 2024 recidivism rate, 22.7 percent, was the second-lowest in state history. And in the most recent data available, the population began ticking back up between 2021 and 2023, driven by people serving longer minimum sentences, even as crime nationally sat near a 30-year low. Sentence length and crime rate are not moving in lockstep in any of these periods. That absence of a relationship is exactly what the deterrence research predicts, and exactly what the Snow doctrine assumes doesn’t happen.

Deterrence Assumes a Calculation That Isn’t Happening

The Snow doctrine treats deterrence as a rational transaction: a person weighs the potential punishment against the potential reward and chooses accordingly. That model has a well-documented blind spot. It assumes the person doing the calculating is operating from a stable baseline, and for a large share of the incarcerated population, that baseline was disrupted long before any sentencing hearing.

Research on adverse childhood experiences, the ACE framework built from a landmark CDC and Kaiser Permanente study, has been applied repeatedly to incarcerated populations with strikingly consistent results. Studies summarized in a 2024 systematic review found more than half of incarcerated people report at least one adverse childhood experience, and separate research comparing incarcerated adults to the general population found men in prison were roughly five times more likely, and women roughly four times more likely, to report four or more such experiences than adults in the original community-based study. Physical abuse, emotional neglect, household substance use, and household incarceration were among the most common.

None of that excuses criminal conduct, and nothing here argues that it should. It does mean the person a deterrence-based sentence is supposedly warning off is disproportionately someone whose relationship to risk, impulse, and authority was shaped by trauma the system never treated. A sentencing doctrine built on rational deterrence has very little to say to someone whose decision-making was never primarily rational to begin with. Longer sentences do not retroactively install the coping mechanisms, treatment, or stability that were absent in the first place.

What “Tough” Sentencing Actually Looks Like on the Record

Jackson County Circuit Judge John McBain is the most publicly documented example of Michigan’s deterrence doctrine pushed to its most visible extreme. In the case of Dawn Marie Dixon-Bey, convicted of second-degree murder, McBain sentenced her to 35 to 70 years, 15 years above the sentencing guidelines, reasoning from facts the jury had specifically declined to convict on. The Michigan Court of Appeals vacated the sentence. McBain defied the order and resentenced her to 30 to 70 years anyway. A second Court of Appeals panel vacated that too, writing that a trial judge unable to follow a higher court’s ruling “is in the wrong line of work,” and suggested the episode could warrant a Judicial Tenure Commission investigation. At an earlier sentencing, McBain told a different defendant, “I hope you die in prison.”

McBain is an outlier in degree, not in kind. Michigan appellate courts have separately upheld top-of-guidelines sentences from other judges, including a 225-month minimum from Lenawee County’s 39th Circuit Court and a similar top-end sentence from a Wayne County judge, over defense arguments that the punishment exceeded what the underlying conduct warranted. Those sentences were not misconduct. They were exactly what the guidelines allow. The point is not that any one judge is uniquely punitive. It is that “send a message” sentencing is normalized across the bench, resting on a deterrent effect the research does not support, whether the judge imposing it draws national attention or none at all.

Michigan’s Report Card on Its Own Deterrence Claim

The Counterargument: “Judges Follow the Law as Written”

That is true, and it matters. Individual judges did not write Snow, and a sentencing judge who cites deterrence is applying binding precedent, not inventing a justification. The criticism here is not that judges are lying when they invoke deterrence. It is that the doctrine itself has not been updated to reflect forty years of research on what actually deters, and that the state has built no functioning mechanism to check whether its own sentencing theory produces the results it claims. A legal requirement to consider a factor is not the same thing as evidence that the factor works the way the law assumes it does.

What This Means for the People Who Run This System

So here is what I want judges, legislators, and MDOC leadership to sit with.

You are legally required to invoke deterrence. You are not legally required to prove it works, and right now, nothing in Michigan is built to make you.

An ombudsman that can’t subpoena and can’t be subpoenaed is not oversight. It’s theater with a budget line.

A FOIA law that excludes the people most affected by the agency it covers is not transparency. It’s a lock with the key on the wrong side of the door.

And four deaths at one facility in two months, following a fifth months earlier, while thirty lawmakers from both parties ask the state’s own corrections director to resign, is not an oversight gap anymore. It’s the predictable output of a system that was never built to answer for itself.

Deterrence is the law. Checking whether it works has never been anyone’s job.