# Patch Notes for Society #001: Prisons

## How We Got Here, and How We Build a Better Prison System

Subtitle: A practical prison reform migration plan for accountability, rehabilitation, restitution, reentry, funding, staffing, legal change, and fewer future victims.

This is part of **Patch Notes for Society**, a series about public systems we inherited, the bugs they create, and the upgrades we should test next.

Start here if you want the core moral argument: [A Better Prison System](https://medium.com/@sethsaps/a-better-prison-system-284caa61c7cb).

This essay is the second layer: history, incentives, where we are today, why the current system persists, and what a realistic migration plan could look like.

In my last essay, I argued that prison should continue to protect society and hold people accountable, but that it should also be redesigned around a clearer goal: reducing the chance that someone creates another victim after release.

That idea raises a harder question.

If the current system is so obviously incomplete, how did we get here? Why does it persist? And if we wanted to build something better, what would the actual migration plan look like?

This essay is an attempt to answer those questions. I am also turning this into a more formal policy memo under **The Better Systems Project**, and I am actively looking for criticism from researchers, corrections leaders, victim advocates, lawyers, funders, and policymakers.

## What Prison Is Supposed To Do

A prison system can serve several purposes at once.

It can punish people for serious wrongdoing. It can separate dangerous people from the public. It can deter future crimes. It can require accountability to victims. It can create conditions for rehabilitation. It can prepare people to return to society in a way that makes everyone safer.

The problem is not that any one of these goals is illegitimate.

The problem is that our system has become much better at confinement than transformation. It can count bodies. It can count sentences. It can count beds. But it is much worse at answering the question that should matter most:

Did this sentence make the public safer over time?

## How We Got Here

The American prison system did not become what it is today by accident.

Over time, it absorbed many different ideas about justice: religious penance, forced labor, racial control, public order, deterrence, rehabilitation, incapacitation, and political toughness.

Some eras emphasized reform. Others emphasized punishment. In the late twentieth century, the political center of gravity shifted heavily toward longer sentences, mandatory minimums, three-strikes laws, aggressive drug enforcement, and reduced tolerance for parole or early release.

Some of this came from real public fear. Crime was high in many communities. Victims were being ignored. Politicians were punished for seeming weak. Prosecutors gained power. Legislatures passed laws that made sentencing more rigid. Prisons became economic anchors in some rural communities. Private companies learned how to profit from the growing system.

The result was not one single prison system. It was a patchwork of federal prisons, state prisons, local jails, probation systems, parole systems, private vendors, courts, prosecutors, sheriffs, and legislatures. Each part has its own incentives. No single actor fully owns the long-term outcome.

That matters because a fragmented system can fail while every individual part claims it is just doing its job.

## Where We Are Today

The United States still relies heavily on incarceration and correctional supervision.

Millions of people move through prisons, jails, probation, and parole. The Bureau of Justice Statistics reported that at the end of 2023, about 5.6 million adults were under correctional supervision in the United States, including about 1.85 million people incarcerated in state or federal prisons or local jails and about 3.77 million people supervised in the community on probation or parole. Local jails alone held about 657,500 people at midyear 2024 and reported 7.9 million admissions from July 2023 through June 2024.

Many people released from prison are arrested again within a few years. BJS followed people released from state prisons in 24 states in 2008 and found that about 66 percent were arrested within 3 years and 82 percent were arrested within 10 years. About 61 percent returned to prison within 10 years for a new sentence or a parole/probation violation.

Those numbers do not mean every person released from prison is dangerous. Arrest is not conviction. Recidivism statistics can be messy. But the basic pattern is hard to ignore: if the goal is fewer future victims, the system is not performing well enough.

The moral problem is obvious. The practical problem is just as serious.

If prison releases people less stable, less employable, more traumatized, more disconnected from family, and still untreated for addiction or mental illness, then prison may satisfy a demand for punishment while increasing the risk of future harm.

That is not accountability. That is delayed failure.

## The Perverse Incentives

The current system has incentives that often point in the wrong direction.

Prisons are commonly funded around capacity, staffing, and operations, not around whether people successfully reintegrate after release. Vendors can make money from phone calls, commissary, food, healthcare, electronic monitoring, and other services whether outcomes improve or not. Some towns depend economically on prisons staying full. Prosecutors are usually rewarded for convictions, not for whether their decisions reduced crime five years later. Politicians are rewarded for sounding tough, especially after high-profile crimes.

Probation and parole can also create traps. Supervision is supposed to support reintegration and protect the public. But when the system is overloaded, it can become a compliance machine that sends people back to jail or prison for technical violations rather than new serious harm.

The deepest incentive problem is that almost nobody is held accountable for the full lifecycle.

If someone leaves prison without housing, employment, treatment, identification, transportation, or a realistic plan, and then commits another crime, the system treats that as the individual's failure. And it often is. But it is also a system failure if we knew those risk factors existed and did little to address them.

## Restitution Is Necessary, But Not Enough

In the last essay, I argued that people who cause harm should be expected to make it up to victims where possible.

I still believe that.

Restitution matters because accountability should be concrete. A system that asks nothing of the person who caused harm is incomplete. Victims should not be erased from reform. They should be central to it.

But restitution is not enough.

Some harm cannot be repaired. Money, apology, service, or structured restitution may help, but they cannot undo murder, assault, trauma, lost years, or permanent injury.

Many victims also do not want contact. Some may want dialogue or restitution. Others may want distance, privacy, safety, or no involvement at all. A serious system has to respect that.

Most importantly, restitution after harm is morally second-best. If we can prevent a future robbery, assault, shooting, overdose, or abuse, that is better than compensating someone afterward.

So the goal should be twofold:

A better prison system should require people to take responsibility for harm already caused, and reduce the chance that anyone else is harmed in the future.

Restitution is how the system responds to past harm. Rehabilitation is how the system prevents future harm. A serious justice system needs both.

## Breaking the Cycle of Harm

There is another uncomfortable part of this conversation.

Many people in prison were harmed long before they harmed someone else. Some grew up around violence, neglect, poverty, addiction, untreated mental illness, unstable housing, failing schools, or communities where the state mostly appeared as police and punishment.

That history matters.

But it does not erase responsibility.

Paulo Freire, in *Pedagogy of the Oppressed*, warned that people who have been oppressed can internalize the logic of the oppressor. In other words, being harmed does not automatically make someone just. Sometimes people who are denied dignity learn to deny dignity to others.

That is one reason prison reform has to reject two false choices.

The first false choice is to say that because someone suffered, they are not responsible for the suffering they caused.

The second false choice is to say that because someone caused suffering, their own suffering no longer matters.

Both are wrong.

A better system should be able to hold two truths at once: people are shaped by what happened to them, and they are responsible for what they do to others.

The goal is not to turn victims into villains or people in prison into passive products of circumstance. The goal is to break the chain. If someone was hurt and then hurt someone else, the justice system should interrupt that pattern, not simply warehouse it until it reappears in another victim's life.

That is why rehabilitation is not an act of sympathy alone. It is a public safety strategy.

And there is evidence that at least some interventions can work. A RAND meta-analysis found that incarcerated people who participated in correctional education programs had 43 percent lower odds of recidivating than those who did not. That does not prove every program works, and it does not mean education alone can fix prison. But it does show that prison does not have to be dead time. Some uses of that time produce better public safety outcomes than others.

## Where We Should Be Going

The next prison system should be built around outcomes, not slogans.

It should still incapacitate people when they pose a serious danger. It should still punish serious wrongdoing. It should still recognize the pain of victims and families. But it should stop pretending that confinement by itself is a complete public safety strategy.

Every sentence should include a public-safety improvement plan.

That plan should ask:

- What risk factors contributed to this crime?
- Which of those factors can realistically be changed?
- What education, treatment, job training, restitution, and supervision are required?
- What would make this person less likely to harm someone after release?
- What does the victim need, and how can the system respect that?
- How will we measure whether the intervention worked?

That is the shift: from an incarceration-centered system to an outcome-centered system.

## A Practical Strategy

The reform cannot happen all at once. A national overnight redesign would fail.

The practical path is a managed migration.

First, choose a pilot. It could be one state, one facility, or one category of incarcerated people. Start where the political environment is possible and where the prison leadership is willing to measure results honestly.

The first pilot should be narrow enough to measure honestly. For example: one medium-security facility, 500 to 1,000 participants, people within 18 to 36 months of release, and exclusions only where there is a clear safety or operational reason. A second pilot could focus on people repeatedly cycling through jail for lower-level offenses, addiction, mental illness, or technical violations.

Second, define the metrics before launching. The system should track recidivism, new violent offenses, employment, education completion, treatment completion, restitution collected, prison safety, victim satisfaction where appropriate, housing stability after release, cost per participant, and cost per successful reentry. The public should see the results, including failures.

Third, build the operating model. That means intake assessments, individualized plans, education tracks, job training, addiction treatment, mental health care, family contact, victim services, restitution accounting, and pre-release planning that begins long before the release date.

Fourth, staff it properly. This system would need educators, therapists, addiction specialists, case managers, correctional officers, victim advocates, job trainers, employers, legal aid, community colleges, healthcare providers, housing partners, and formerly incarcerated mentors who can help people navigate reentry.

Fifth, create the supply chain. A rehabilitative prison system depends on reliable partners: trade schools, employers, credentialing bodies, transportation providers, treatment clinics, housing organizations, technology vendors, and local nonprofits.

Sixth, change the funding model. Money should move toward programs that reduce future harm. Savings from lower reincarceration should be reinvested into reentry, victim services, education, and treatment. Grants should reward measurable outcomes, not just program activity.

The initial funding model should not depend on vague optimism about future savings. It should have a real budget. A practical pilot would likely need dedicated funding for staff, programming space, technology, evaluation, victim services, and post-release support. A reasonable structure would be: 40 percent for human capital, 20 percent for education and job training, 15 percent for behavioral health and addiction treatment, 10 percent for victim services and restitution administration, 10 percent for housing/employment transition support, and 5 percent for independent evaluation and data infrastructure.

Over time, the model should be funded partly by reinvesting avoided reincarceration costs. But the savings should not simply disappear into a general budget. They should be locked into a public safety reinvestment fund: victim services, reentry housing, treatment capacity, prison education, and community supervision that actually supports reintegration.

Seventh, change the law gradually. Some laws should be removed or narrowed: excessive mandatory minimums, some three-strikes provisions, unnecessary licensing bans, excessive fines and fees, and reincarceration for low-level technical violations. Other laws should be added: earned-time credits for verified progress, independent prison oversight, guaranteed reentry planning, prison education access, data transparency, and outcome-based funding.

The key is sequencing. Do not change every sentence, facility, and statute at once. Prove what works. Expand what works. Stop funding what does not.

## The Hard Objections

Any serious reform has to answer its strongest critics.

The conservative objection is that this could become a soft-on-crime project that risks public safety. The answer is that the goal is not less accountability. The goal is better accountability. A system that releases people more likely to commit new crimes is not tough. It is careless.

The prosecutor's objection is that some people are dangerous and some crimes deserve severe punishment. That is true. A better prison system still needs incapacitation. Some people should be separated from the public for a long time. But even long sentences should be used to reduce future harm where possible.

The correctional officer's objection is that reformers often underestimate facility safety, staffing shortages, gangs, contraband, mental illness, and day-to-day operational reality. That objection should be taken seriously. Staff should be part of pilot design, not treated as obstacles. If the plan does not make facilities safer for staff and incarcerated people, it is not good enough.

The victim advocate's objection is that reform can erase victims by focusing too much on people who caused harm. That is why restitution, trauma support, victim voice, and safety planning have to be built into the model from the beginning. Rehabilitation should not compete with victim services. It should help prevent the next person from becoming a victim.

The budget objection is that this sounds expensive. It is expensive. But the current system is also expensive, and it often buys poor outcomes. The right comparison is not reform versus free. It is reform versus continuing to pay for incarceration, reincarceration, victimization, emergency healthcare, homelessness, unemployment, and supervision failure.

## A Ten-Year Migration Timeline

The first 6 months should be about baseline and design. Pick the pilot. Audit current costs, staffing, recidivism, programming, violence, education, healthcare, and reentry outcomes. Define the success metrics. Identify the laws and policies that block reform. Build a coalition that includes corrections leaders, victims' advocates, prosecutors, defense attorneys, judges, employers, educators, and formerly incarcerated people.

From 6 to 18 months, launch the pilot. Start with a limited population. Add education, job training, therapy, addiction treatment, restitution planning, and reentry case management. Train staff for the new operating model. Establish independent oversight and public reporting.

From 18 to 36 months, evaluate and expand. Compare the pilot against the baseline. Expand only the programs with measurable results. Start legislative reforms around earned-time credits, technical violation limits, licensing reform, employment barriers, and reentry funding.

From years 3 to 5, institutionalize the model. Move from one pilot to multiple facilities. Create permanent budget lines. Scale community college, employer, treatment, and housing partnerships. Replace occupancy-based incentives with outcome-based funding. Publish annual public safety and reentry scorecards.

From years 5 to 7, migrate the broader system. Apply the model to most eligible incarcerated people. Reduce reincarceration for technical violations. Shift savings from lower prison churn into victim services, treatment, education, and reentry. Expand sentencing, parole, expungement, and collateral-consequence reforms.

From years 7 to 10, make the new model the default. Rehabilitation, restitution, and reentry planning should no longer be experimental. Every sentence should include a public-safety improvement plan. Funding should follow outcomes: fewer victims, fewer returns to prison, more restitution, more employment, and safer facilities.

## What Could Go Wrong

This approach could fail.

Programs can become performative. Metrics can be gamed. Politicians can abandon reform after one high-profile crime. Agencies can cherry-pick low-risk participants to make numbers look better. Staff can resist changes they see as naive or unsafe. Vendors can rebrand old services as reform. Advocates can understate the seriousness of violent crime. Opponents can pretend that any rehabilitation is disrespectful to victims.

That is why transparency matters.

A better system should be compassionate, but not vague. It should be humane, but not soft-headed. It should be evidence-based, independently audited, and willing to shut down programs that do not work.

## What Each Person Can Do

If you think this framing is directionally right, do not just share it as an opinion. Turn it into pressure, review, or implementation.

If you are a voter, ask candidates what they will do to reduce future victimization. Do not settle for slogans about being tough or soft on crime. Ask how they measure public safety over time.

If you contact your member of Congress, ask them to support prison education, job training, earned-time credits for verified progress, independent oversight, and federal grants tied to lower recidivism and stronger reentry outcomes.

If you contact your state legislators, focus on sentencing reform, parole and probation reform, reentry funding, licensing reform for people with records, prison education access, and transparent reporting on outcomes.

If you contact your governor, ask them to appoint corrections leaders who are measured by results, expand evidence-based reentry programs, require public prison outcome reporting, and make corrections reform a public safety priority.

If you contact your county prosecutor or district attorney, ask whether they track outcomes after conviction. Ask whether they support diversion, treatment, restitution, and supervision models when those approaches are more likely to prevent future harm.

If you contact your sheriff or local jail officials, ask what reentry planning begins before release. Ask whether people are connected to housing, medication, identification, treatment, and employment help before they leave custody.

If you are an employer, create a fair-chance hiring pathway. Partner with reentry organizations. Hire based on skill, reliability, and risk relevance rather than blanket exclusions.

If you are a lawyer, help remove legal barriers to reentry. Support expungement and record-sealing clinics. Challenge excessive fines, fees, and technical violations. Help translate reform ideas into statutory language.

If you are an educator, partner with prisons, jails, or reentry nonprofits. Offer credentialed courses. Build pathways from prison education into community college, trade school, or employment.

If you are a victim advocate, insist that reform include victims rather than erase them. Push for restitution, trauma services, safety planning, and victim voice.

If you are a researcher, corrections leader, policymaker, funder, prosecutor, public defender, reentry operator, or prison staff member, I would value critique more than agreement. The policy memo version is being built around evidence, implementation constraints, funding, legal changes, and failure modes. Tell me what is wrong, naive, missing, or overstated.

If you want the broader series, follow future **Patch Notes for Society** issues. The next topics on the list include housing, healthcare affordability, education, addiction and mental health, money in politics, immigration, climate and energy, cost of living, and gun violence.

## The Real Question

The question is not whether we care more about victims or people in prison.

The question is whether we are willing to build a system that creates fewer victims in the first place.

That requires accountability for the person who caused harm. It also requires accountability for the system that claims to protect the public.

If prison is going to take years of a person's life in the name of justice, it should use those years to do more than punish. It should use them to reduce future harm.

That is not idealism.

That is what public safety should mean.

## What Comes Next

This essay is not the final word. It is the public version of a working proposal.

The next step is expert review:

- correctional researchers: does the evidence support the model?
- victim advocates: does the framing respect victims and safety?
- corrections operators: what would break inside real facilities?
- policymakers: what laws and funding mechanisms would actually need to change?
- voters: which leaders are willing to measure public safety by outcomes rather than slogans?

If the feedback improves the proposal, I will revise it publicly.

That is the point of Patch Notes for Society: not to pretend the first draft is right, but to make public systems easier to inspect, criticize, and improve.

## Sources

- Bureau of Justice Statistics, *Correctional Populations in the United States, 2023*: https://bjs.ojp.gov/library/publications/correctional-populations-united-states-2023-statistical-tables
- Bureau of Justice Statistics, *Jails Report Series: 2024 Preliminary Data Release*: https://bjs.ojp.gov/library/publications/jails-report-series-2024-preliminary-data-release
- Bureau of Justice Statistics, *Recidivism of Prisoners Released in 24 States in 2008: A 10-Year Follow-Up Period*: https://bjs.ojp.gov/library/publications/recidivism-prisoners-released-24-states-2008-10-year-follow-period-2008-2018
- RAND Corporation, *Evaluating the Effectiveness of Correctional Education*: https://www.rand.org/pubs/research_reports/RR266.html