When possibility becomes enough to justify permanent control
Day 5 of 43 in The Punishment Machine
There is one word powerful enough to extend punishment indefinitely.
Risk.
A person completes prison.
Risk remains.
A person follows every rule for ten years.
Risk remains.
A person grows older, becomes ill, completes treatment, maintains housing, supports a family, and commits no new offense.
Risk remains.
The word does not require proof that someone has done something wrong today.
It points toward what might happen tomorrow.
That makes risk different from nearly every other basis for punishment.
A crime is an act.
Risk is a prediction.
A sentence is supposed to respond to what a person did.
Risk permits the system to continue restricting someone because of what officials fear the person might someday do.
That fear may be sincere. It may even be reasonable in some cases.
But once risk becomes the answer to every request for relief, punishment no longer needs an ending.
There is always some possibility of future harm.
No human being has a zero-percent chance of doing something wrong.
If zero risk is the standard for freedom, no one can ever satisfy it.
The Safest Word in Government
Risk is one of the safest words an official can use.
It sounds scientific.
It sounds responsible.
It suggests caution rather than punishment.
A judge who continues supervision because of risk appears protective.
An agency that denies relief because of risk appears prudent.
An officer who recommends more restrictions because of risk appears careful.
The word also protects the decision-maker.
Granting relief creates a visible danger. If the person later commits another offense, someone may ask why the restriction was removed.
Continuing punishment creates a largely invisible harm. The person may lose years of freedom, employment, travel, privacy, and family life, but the official who refused relief is unlikely to face criticism.
This creates a powerful imbalance.
The cost of releasing someone who later fails can become public.
The cost of continuing unnecessary restrictions usually remains private.
As a result, the system has every incentive to overestimate danger and very little incentive to recognize when risk has declined.
The safer institutional answer is almost always:
Not yet.
Continue supervision.
Keep the restriction.
Deny the request.
Reconsider later.
Later can become forever.
Risk Is Not a Single Thing
The word risk often hides more than it explains.
Risk of what?
A technical violation?
Drug use?
Failure to report?
A minor offense?
Violence?
Serious harm?
Any contact with the criminal justice system?
These are not equivalent outcomes.
A person may present a modest risk of missing an appointment while presenting almost no risk of violence.
Another person may have a history of unstable housing but no history of harming anyone.
A risk classification that collapses all undesirable outcomes into one category can make a person appear more dangerous than the underlying facts justify.
There is also the question of time.
Risk over the next month is not the same as risk over the next twenty years.
A statement that someone has a chance of reoffending eventually may sound alarming even when the annual likelihood is very small.
Numbers can be presented in ways that frighten rather than inform.
The longer the time horizon, the larger a cumulative percentage may appear.
But public policy should not be driven by the most dramatic way of expressing uncertainty.
It should ask what the actual probability is, what kind of harm is being predicted, and whether the restriction imposed is proportionate to that danger.
Without those distinctions, risk becomes less an assessment than an incantation.
Say the word, and the discussion ends.
Static Facts and Living People
Many risk judgments rely heavily on static factors.
The nature of the original offense.
Age at the time of the offense.
Criminal history.
The number of prior arrests or convictions.
The sentence imposed.
These facts may matter.
But they never change.
A person cannot alter the age at which something happened.
He cannot rewrite the conviction.
He cannot make the original charge less serious.
If those static facts dominate every later decision, then rehabilitation becomes almost irrelevant.
The system may tell people to complete treatment, obey conditions, maintain employment, repair family relationships, and live lawfully.
But when they later request relief, the government may return to the same facts that existed on the day of sentencing.
The message is contradictory:
Change your life.
But understand that your changed life may never outweigh your past.
A fair assessment must consider dynamic facts as well.
What has the person done since the offense?
How long has he remained law-abiding?
Has he completed treatment?
Has he accepted responsibility?
Has he maintained stable housing?
Has he supported himself and others?
Has he complied with supervision?
Has his health changed?
Has his age reduced the likelihood of future offending?
Has he demonstrated self-control in the real world, outside an institution?
These facts describe the person who exists now.
Static factors describe the person and circumstances that existed then.
Justice requires both.
The Problem With Risk Scores
Risk-assessment tools promise consistency.
Instead of relying only on instinct, officials can use data to estimate the likelihood of future misconduct.
That can be better than pure guesswork.
Human judgment is vulnerable to fear, bias, mood, politics, and personal experience. A structured assessment may force decision-makers to consider factors they might otherwise overlook.
But a score is not a prophecy.
It is an estimate based on patterns found in groups.
It does not know the future of the individual standing before the court.
A person may share characteristics with a group that has a particular average outcome. That does not mean he will personally produce that outcome.
Statistics describe populations more reliably than they predict individual lives.
The distinction is easy to forget when a score is printed on an official form.
Numbers create an appearance of precision.
A person becomes low, medium, or high risk.
The category then begins to function like an identity.
High risk.
Not a person assessed as presenting an elevated statistical probability under a particular model at a particular moment.
Just high risk.
The label can follow someone from one decision to the next, even when the underlying information becomes outdated.
A score may also include factors closely connected to poverty, neighborhood, education, employment history, family structure, or prior contact with the justice system.
Those factors may improve prediction while also reinforcing existing disadvantage.
A person can be classified as riskier because he has fewer resources, then denied opportunities because he is classified as risky.
The score predicts instability.
The system responds by making stability harder to achieve.
Perfect Compliance Can Disappear
Years of successful conduct should be powerful evidence.
Yet compliance is often treated as merely what the person was required to do.
A judge may say that following the rules is expected and therefore does not justify relief.
At one level, that is true.
People on supervision are supposed to comply.
But if compliance does not count as evidence of reduced risk, what does?
A person cannot prove that he will never commit a future offense.
He can only demonstrate his judgment through conduct over time.
Every year without a violation is real-world evidence.
Every approved trip completed without incident is evidence.
Every year of employment is evidence.
Every stable relationship is evidence.
Every problem handled without criminal conduct is evidence.
Compliance should not automatically entitle someone to relief.
But neither should it be dismissed as meaningless.
When the government predicts future behavior, observed behavior is among the best evidence available.
A system that relies on the original crime while minimizing years of successful conduct is not truly assessing risk.
It is preserving punishment.
Age Changes Risk
People change with age.
Energy changes.
Health changes.
Priorities change.
Opportunity changes.
Impulsivity often declines.
The person standing before a court at fifty, sixty, or seventy is not identical to the person who committed an offense decades earlier.
This does not erase responsibility.
It does affect prediction.
A risk assessment that fails to account meaningfully for aging may exaggerate danger by treating the past as though it remains biologically and socially unchanged.
Health can matter as well.
Serious illness, disability, reduced mobility, cognitive change, and dependence upon caregivers may alter both opportunity and capacity.
These considerations should not become automatic excuses.
They should become part of an honest evaluation.
The question is not whether the original conduct was serious.
The question is whether the current person presents enough present danger to justify continued control.
Those are different inquiries.
Risk Can Never Be Eliminated
Every act of freedom involves risk.
People drive cars despite the risk of accidents.
They raise children despite the risk of neglect or abuse.
They operate businesses despite the risk of fraud.
They drink alcohol despite the risk of addiction or violence.
They own tools, use the internet, travel, marry, divorce, work, vote, and make private decisions despite the possibility of harm.
Ordinary citizenship does not require proof of harmlessness.
People with criminal convictions are often held to a different standard.
They may be asked, directly or indirectly, to prove that nothing bad could ever happen.
That is impossible.
Government should manage meaningful risk, not demand imaginary certainty.
The appropriate question is not:
Can we guarantee that this person will never fail?
The appropriate questions are:
How likely is serious harm?
What evidence supports that conclusion?
Has the risk changed?
Does the restriction reduce that risk?
Is the burden proportionate?
Could a narrower measure protect the public?
How often will the decision be reviewed?
What must the person do to earn relief?
Without clear answers, risk becomes a permanent veto.
The Difference Between Caution and Cowardice
Caution is necessary in criminal justice.
Decisions affect victims, families, communities, and public safety.
But caution requires more than preserving every restriction.
It requires the courage to distinguish between people.
To recognize change.
To admit that an old prediction may no longer fit.
To remove a condition when it no longer serves a purpose.
To accept that no decision can eliminate uncertainty.
A system that never releases control is not necessarily wise.
It may simply be afraid to take responsibility for ending punishment.
There should be no shame in saying that a person remains dangerous when evidence supports that conclusion.
There should also be no shame in saying that a person has changed enough that extraordinary restrictions are no longer justified.
Both decisions require judgment.
Only one is routinely treated as politically safe.
Risk Must Have a Burden of Proof
When risk is used to continue punishment, the government should have to explain itself.
Not with general statements about the seriousness of crime.
Not with the bare fact of conviction.
Not with speculation that something might happen.
The explanation should identify the present concern, the evidence supporting it, the relationship between the concern and the restriction, and the reason lesser measures are inadequate.
The greater the burden on liberty, the stronger that explanation should be.
A lifetime condition should require more than a lifetime label.
A denial of relief should identify what the person has failed to demonstrate.
A review process should provide a realistic path forward.
Otherwise, the person is trapped in an impossible assignment:
Prove the absence of a future event.
No one can do that.
Risk is real.
So is change.
A fair system must be capable of recognizing both.
Tomorrow, we close the first week by examining what life looks like when freedom is no longer assumed, but granted one decision at a time.
When every door requires approval, freedom becomes permission.
