A sentence measures time. Justice must decide what that time has accomplished.
Day 31 of 43 in The Punishment Machine
The judge announces twenty years. Thirty years. Life. The number sounds precise.
Time is simple on a calendar. Justice is not. Twenty years is twenty years, but the person at the end of those years may not be the person who began them.
It is not.
The court sees the offense, the victim, the evidence, and the defendant standing before it. It cannot see the sixty-five-year-old who may someday occupy the body of the twenty-five-year-old being sentenced. It cannot know whether he will remain dangerous, change, become disabled, find remorse, or grow unrecognizable even to his younger self.
A long sentence is more than punishment for the past. It is a prediction about the future. It assumes that the reasons for confinement will remain strong enough, year after year, to justify every day imposed at the beginning. Sometimes that prediction will prove correct.
Sometimes it will not.
The difficult question is whether the justice system should ever be required to look again.
What Is the Sentence Supposed to Accomplish?
Punishment is usually defended through several purposes:
Retribution.
Deterrence. Incapacitation. Rehabilitation. Respect for the law.
Protection of the public.
Recognition of the harm done to victims. These purposes may justify a substantial sentence. Serious offenses may deserve serious punishment, dangerous people may require confinement, and victims may properly expect moral condemnation. But those purposes do not necessarily remain equally strong forever. Retribution asks what punishment the offense deserves.
Incapacitation asks whether the person must remain confined to prevent harm. Deterrence asks whether continued punishment discourages this person or others. Rehabilitation asks whether the person has changed. A sentence may continue serving one purpose after another has weakened. The punishment may remain deserved even when continued confinement is no longer necessary for safety.
The question is not whether the offense has become less serious. It is whether another ten years in prison will accomplish enough to justify taking another ten years of life.
Time Is Not Empty
A long sentence can be discussed as though year one, year ten, and year twenty-five were identical units.
They are not.
Time changes people.
A young person may mature.
An impulsive person may develop control.
An addicted person may recover.
A healthy person may become disabled. A parent may become a grandparent. Someone who once presented substantial danger may age beyond much of that danger. The United States Sentencing Commission found a strong relationship between age and recidivism among federal offenders. Over an eight-year follow-up period, 13.4 percent of people released at age sixty-five or older were rearrested, compared with 67.6 percent of those released before age twenty-one. Recidivism declined with age whether measured by rearrest, reconviction, or reincarceration. (U.S. Sentencing Commission)
That does not mean every older prisoner is safe. Age does not erase the offense, a history of violence, institutional misconduct, threats, or other evidence of danger. It does mean the passage of time is relevant. A person sentenced at twenty-five should not automatically be judged at sixty-five as though nothing except the calendar has changed.
Longer Is Not a Simple Science
Research does not support the easy claim that every additional year in prison creates a predictable increase in public safety. A 2022 Sentencing Commission study found lower odds of recidivism among matched federal offenders who served more than sixty months, with the largest difference among those sentenced to more than 120 months. For sentences of sixty months or less, the study found no statistically significant difference. A separate Commission analysis of aging found no clear relationship between sentence length and rearrest among people sentenced to at least one year, while age remained strongly associated with declining recidivism. (U.S. Sentencing Commission)
These findings should encourage humility.
A prison sentence is not a medication for which science has established one correct dosage. Twenty years does not guarantee twice the deterrence of ten. Forty years does not automatically create twice the public safety of twenty. The effects depend upon the person, offense, conditions of confinement, age at release, and available support.
The number imposed at sentencing remains a judgment. It should not be mistaken for scientific certainty.
The Person Who Was Sentenced May No Longer Be the Whole Person
The person remains legally and morally connected to the offense. He cannot escape responsibility by claiming to have become someone else. But identity is neither perfectly fixed nor completely replaced. The person who committed the crime remains part of the person who exists decades later.
He may no longer be the whole of that person. Some prisoners spend more years confined than they ever lived free as adults. Some become better, some worse, and some merely learn to appear compliant. Others develop remorse and insight that could not have been forced at sentencing.
The justice system should not presume transformation. It should be capable of examining it.
Good Conduct Is Evidence, Not Proof
A prisoner seeking reconsideration may present years without violations, education, treatment, work, mentoring, remorse, staff support, and a release plan. The system may respond that good conduct is expected.
That is partly true.
Following prison rules does not prove someone can live safely in the community. But if decades of conduct count for almost nothing, what evidence of change could ever be enough? A prisoner cannot demonstrate community success while the government prevents him from entering the community. He can offer only the evidence available inside the institution.
That evidence should be examined carefully—neither accepted uncritically nor dismissed as routine. A spotless record does not guarantee transformation. A long pattern of responsible conduct remains more informative than pretending the original offense is the only meaningful act of the person’s life.
Finality Should Protect Judgments, Not Freeze People
Sentences cannot be reopened every few months. Victims deserve stability, families need certainty, and courts cannot endlessly relitigate the same facts. Witnesses disappear and evidence deteriorates. The law has a legitimate interest in finality. But finality should mean that the original conviction and sentence cannot be attacked endlessly without legal grounds.
It should not require pretending that decades of human life produce no new facts. A second-look process need not ask:
Was the sentencing judge wrong?
It can ask:
Does the sentence remain appropriate for the person and circumstances that exist now?
The first question challenges the past. The second recognizes that time has created new evidence.
Reconsideration Is Not Automatic Release
A meaningful review would not promise freedom.
It would promise judgment.
The prisoner would bear the burden of showing change, present circumstances, current risk, and a workable release plan. The government could present the seriousness of the offense, institutional conduct, unresolved risks, victim concerns, and reasons confinement remains necessary. Victims should receive notice and an opportunity to be heard.
The court could examine the original harm, the person’s role, time served, age, conduct, treatment, remorse, health, continuing danger, support, release plans, and whether a less restrictive measure could protect the public.
The answer might remain no.
For some people, it should remain no. But the no should come from a current individualized judgment—not merely from the number imposed decades earlier.
The Existing Door Is Narrow
Federal compassionate-release law permits courts to reduce an imposed sentence for extraordinary and compelling reasons after considering the federal sentencing factors and applicable policy. Since the First Step Act, qualifying incarcerated people may bring motions themselves after satisfying the statutory administrative requirement. That is not a broad second-look system.
It is an exceptional remedy.
In fiscal year 2025, the Sentencing Commission reported 394 grants among 2,795 compassionate-release motions with determinable outcomes—14.1 percent. (U.S. Sentencing Commission) Compassionate release is important, especially for severe illness, aging, family circumstances, or other extraordinary conditions. But “How long is long enough?” is a broader question. A healthy person may have changed.
A person may no longer present the risk anticipated decades earlier. A sentence may have fulfilled its legitimate purposes without a terminal illness or medical crisis. A true second-look process would examine justice after substantial time—not merely emergency.
The Victim’s Time Matters Too
A prisoner may change while the victim continues living with the harm. A person who was killed does not return.
An injury may remain.
Trauma may continue.
A family may experience every release request as an attempt to minimize what happened. Those concerns must not be dismissed. The victim is part of the truth. But victims do not all agree about punishment, release, forgiveness, or mercy. Another decade in prison may feel essential to one and meaningless to another.
The state should listen without transferring the entire responsibility for punishment to the person harmed. The final decision remains a public judgment about justice and safety.
A Sentence Should Be Reviewable Because People Are Not Numbers
A fair second-look system could begin after a substantial period—perhaps ten, fifteen, or twenty years, depending upon the offense and sentence. Review would not need to occur constantly. A denial could establish a reasonable period before another application. Frivolous or repetitive requests could be limited. The process should focus on the evidence created by time:
Who has this person become?
What risk exists now?
What has confinement accomplished?
What purpose would continued imprisonment serve?
Could that purpose be met through supervision, home confinement, treatment, or another less restrictive response? The decision should be written and explained. If release is denied, the court should identify why and what evidence could matter later. If release is granted, the order should include a serious reentry plan and conditions proportionate to the remaining risk.
How Long Is Long Enough?
There is no universal answer.
Ten years may be too long in one case and too short in another.
Age alone cannot decide.
Good conduct alone cannot decide.
The seriousness of the offense alone should not decide every question forever. The answer must come from the purposes of punishment. Long enough is the point at which additional confinement no longer produces enough legitimate justice or public safety to justify its human and financial cost. That point may never arrive for some prisoners.
For others, it may arrive long before the sentence ends. The moral failure is not choosing one answer or the other.
It is refusing to ask.
A sentence records the judgment of one moment. A second look recognizes that human beings continue living after the judgment is entered.
Reconsideration is not forgiveness by stopwatch. It is judgment applied again after time has supplied new evidence.
A sentence should be serious enough to mean something. Justice should also be serious enough to ask, after years or decades, what that sentence has accomplished and what continued punishment is now for.
