Second Chances: What Happens When Prisons Become Classrooms

Start with the population, because it explains most of what follows. People in prison are drawn, overwhelmingly and everywhere, from the group that the school system failed first. Surveys of prison populations across many countries return the same profile: rates of exclusion from school far above the general population, high rates of undiagnosed learning difficulties, literacy levels well below average, and a large share who left education before completing it. Many were in state care as children. Many have head injuries. A great many were, by any reasonable account, already outside the system long before they were inside this one.

This is not offered as an excuse. It is offered as a description of who is actually in the room when a prison education class begins — because the policy question is usually framed as whether criminals deserve free education, and that framing obscures that the people concerned are, in the main, receiving for the first time what they were entitled to at fourteen.

The finding

Research on prison education has been conducted for decades in many jurisdictions, using methods of varying rigour. Synthesising it produces one of the more consistent results in criminal justice research: people who participate in education while incarcerated are substantially less likely to be reconvicted after release than those who do not.

Reviews across the literature typically find reductions in recidivism on the order of a third relative to non-participants, with larger effects for higher levels of education — degree-level study shows the strongest association. Employment after release is also markedly higher among participants.

The obvious objection applies with full force: people who volunteer for education in prison are different from people who don't. They may be more motivated, better supported, further from the offending behaviour, or simply serving sentences long enough to complete a course. Selection could account for a good deal of the difference.

Researchers have worked hard on this — comparing participants to matched non-participants, exploiting waiting lists and capacity limits, and studying what happens when programmes are abruptly cut or restored. The effect shrinks under these treatments. It does not vanish. The considered position in the field is that a meaningful causal component exists, that its size is uncertain, and that the intervention is cheap enough for the cost-benefit case to survive even conservative assumptions.

Why it works — three mechanisms

The recidivism finding is only interesting if we can say why. Three mechanisms are usually proposed, and they are not mutually exclusive.

Employability

The simplest account. A criminal record is a severe employment penalty; a qualification partially offsets it. Someone released with a trade certificate, a literacy level that allows them to complete an application, or a degree has more options than someone released with nothing, and options reduce the pull of the only work they know.

This mechanism is real but limited. It depends entirely on whether employers will hire, and on whether the qualification is recognised outside. Training people in a trade they are then legally barred from entering — which happens, through occupational licensing rules that exclude people with convictions — produces a qualification and no job.

Identity

The mechanism that people who have been through these programmes describe most often, and that criminological research on desistance treats as central.

Leaving offending behind is not primarily a matter of deterrence or opportunity. It involves a shift in how a person understands themselves — from someone who is a criminal to someone who committed crimes and is now something else. That shift needs material. It needs a new description a person can plausibly apply to themselves, and evidence to support it.

Education supplies both. "I am a student" is one of the few socially respectable identities available inside a prison. Completing a course produces external, dated, verifiable proof that the person can do something difficult — which for someone whose record of achievement is otherwise empty is not a small thing. The certificate matters less as a credential than as evidence in an internal argument.

Cognitive and social skill

Structured education develops capacities that are relevant regardless of subject: sustaining attention, tolerating frustration, taking criticism without escalation, considering consequences before acting, and understanding a perspective other than one's own. These are precisely the capacities whose absence features in a great many offending histories. A literature class in which people argue about a character's motives is doing more work than its syllabus suggests.

What it looks like from inside

Descriptions from teachers who work in prisons converge on a few observations that rarely make it into policy documents.

The first is intensity. Prison classrooms frequently contain a level of engagement that teachers elsewhere find remarkable — because participation is voluntary, because the alternative is a cell, and because for many students this is the first classroom they have entered as an adult by choice rather than compulsion. The disciplinary problems that dominate school teaching are largely absent; nobody is there who does not want to be.

The second is the near-universal statement, made in some form by almost every student, that school did not work for them the first time. Often the account includes a specific memory of being told they were stupid, or being moved so often that continuity was impossible, or a difficulty nobody identified. Teachers describe a large amount of early work being spent simply on the proposition that the person is capable of learning, against a lifetime of contrary evidence.

The third is fragility. Transfers between facilities interrupt courses without notice. Lockdowns cancel classes. Access to computers and to the internet is heavily restricted, which makes distance study extraordinarily difficult in an era when almost all materials assume connectivity. Funding for these programmes is frequently the first thing cut and the last thing restored.

The objection, taken seriously

There is a moral objection that deserves a direct answer rather than a dismissal. It runs: law-abiding people pay for their own education, sometimes at great cost. Why should someone who has harmed others receive it free? Is that not a reward for offending, and an insult to the people they harmed?

The feeling behind this is legitimate and should not be waved away. Three responses are worth putting alongside it.

First, on purpose. Almost everyone in prison is coming out. The question is not whether they are released but in what condition. A policy that makes release safer is not a favour to the prisoner; it is a protection for whoever lives on the street they return to. The strongest argument for prison education is made on behalf of future victims who will now not be victims.

Second, on the ledger. If the interest is in what taxpayers pay, prison education reduces the bill. Every prevented return to custody saves far more than the course cost. Cutting it is not a saving; it is a deferred and larger expense.

Third, on what is being given. Basic education is not a luxury being handed out. In most of these cases it is an entitlement that was never delivered — to a child who was excluded at thirteen, whose dyslexia was never identified, who moved through six schools in four years. Providing it at thirty is a late correction, not a gift.

What makes programmes work

The design differences between effective and ineffective provision are consistent enough to list.

• Continuity through transfer and release. The single largest cause of non-completion is being moved. Systems that let a course follow a person — and continue it after release — dramatically raise completion.
• Qualifications that mean something outside. The certificate must be the same one awarded on the outside, without a marking that identifies where it was earned.
• A real ladder. Basic literacy leading to secondary qualifications leading to vocational or degree-level study. Provision that stops at the lowest level tells people how far they are expected to go.
• Vocational training tied to actual hiring. Trades where a record is not an automatic bar, with employers involved in advance.
• Controlled digital access. Secure, restricted learning networks make modern study possible without compromising security. Blanket prohibition is simpler and much more damaging.
• Teachers who are teachers. Not security staff with a syllabus. Programmes staffed by educators who choose the work, and who are trained for the setting, produce visibly different results.
• Release planning that starts early. Education is wasted if a person leaves with a qualification, no accommodation, no identity documents, and no bank account.

The wider frame

There is a way of reading prison education that makes it look like a marginal, slightly sentimental corner of policy. There is another reading in which it is a stress test for the entire proposition this collection is built on.

If education changes lives, it should show up most visibly in the hardest case: adults with the worst educational histories, the fewest resources, the most disrupted circumstances, and the strongest reasons to have given up on the whole idea. That it does show up there — in reconviction rates, in employment, and in the accounts people give of themselves — is the most demanding evidence available that the claim is about something real rather than about selection.

A society decides, one budget line at a time, whether a person's worst act is the last thing that will ever be true about them. Education in prison is one of the few mechanisms by which that question gets answered in the negative — not as forgiveness, and not as leniency, but as the cheapest available method of ensuring that the person who walks out is different from the one who walked in.