Show the Working

Prioritise restorative justice and rehabilitation over short prison sentences

Green · what the evidence says

An independent, source-checked look at Green’s policy “Prioritise restorative justice and rehabilitation over short prison sentences” — what it would actually do across the things that affect your life. Every claim below quotes the source behind it. How this works.

Personal liberty & free speech — Helps

moderate · moderate confidence

By legislating a presumption against short prison sentences and expanding diversion, the policy would reduce the number of people subject to custodial detention — the most severe form of state coercion — replacing it with community-based alternatives that impose lighter obligations. The main caveat is that community sentences and mandatory diversion programmes still carry state-imposed conditions, and the net liberty gain depends on how those alternatives are designed and resourced.

The evidence

Biggest unknown: Whether community sentences and diversion programmes are structured as genuinely voluntary or light-touch, or instead impose burdensome surveillance and compliance conditions that offset the liberty gain from avoiding custody.

Our reading: O10 is concerned with freedom from state coercion — and custodial imprisonment is the most coercive instrument the state regularly deploys against individuals. A legislated presumption against sentences under two years would directly reduce the number of people detained against their will by the state, affecting a population of roughly 49,300 adults annually who currently receive sentences of 12 months or less. Extending the presumption to two years would go further than the existing March 2026 legislation which already mandates suspension of sentences up to 12 months. Community sentences and diversion programmes are not coercion-free — they involve conditions, supervision and compliance requirements — but they impose materially lighter constraints on bodily liberty than incarceration. The net direction for O10 is therefore positive: replacing custody with community alternatives reduces the intensity of state coercion experienced by this cohort. The diversion of drug and alcohol offenders away from criminalisation also reduces exposure to criminal records and associated state-imposed restrictions on life choices. The main liberty caveat is that mandatory diversion or community programmes can themselves carry conditions (drug treatment orders, curfews, reporting requirements) that the evidence provided does not fully characterise. The biggest implementation risk is the fragile state of probation services — 21% officer vacancies and adequate risk assessment in only 28% of cases — which could mean community supervision is either too thin to be meaningful or, if bolstered, more intrusive. That uncertainty keeps confidence at moderate rather than high, but it does not reverse the direction: even an imperfectly resourced community sentence is less coercive than incarceration. The verdict is a moderate improvement to O10.

Public finances & the next generation — Helps

minor · moderate confidence

Replacing short prison sentences with community alternatives and restorative justice is likely to reduce net public spending, because custody costs far more than community orders and reoffending imposes large fiscal costs. The main caveat is that the policy requires upfront investment in an already under-resourced probation service, and if that investment is unfunded or the service cannot absorb demand, savings may not materialise.

The evidence

Biggest unknown: Whether the required investment in probation and community services is funded from savings or requires additional borrowing, and whether a probation service with a 21% vacancy rate can absorb the increased caseload without deteriorating outcomes.

Our reading: The fiscal case for shifting away from short custodial sentences is well-evidenced. Custody costs roughly £47,000 per prisoner per year; community orders cost substantially less. Short sentences drive enormous churn — 49,300 adults cycling through prison annually — and are associated with reoffending rates roughly 20-30 percentage points higher than community alternatives, generating an estimated £20.9bn annual reoffending cost. Restorative justice programmes show a 9:1 savings-to-cost ratio in the evidence. On these numbers, a sustained shift toward community sentences and restorative justice would reduce both direct custodial expenditure and the downstream fiscal costs of reoffending — an improvement to O12. However, the magnitude is constrained by two factors. First, the policy requires 'investment in probation and prison services' without specifying a funding mechanism — if this investment is unfunded, there is a near-term fiscal cost before savings materialise. Second, and more fundamentally, the probation service is already under severe strain: 21% officer vacancies, risk assessments adequate in only 28% of cases, and serious further offences on probation up 55% since 2021-22. Expanding community sentences into this fragile system could reduce the quality of supervision, erode reoffending savings, and increase fiscal risk rather than reduce it. The Prison Reform Trust explicitly warns that 'an unresourced system risks undermining rehabilitation and public safety'. The policy's direction for this parliament has been partly pre-empted by the March 2026 Sentencing Act, which already mandates suspension of sentences under 12 months. The marginal fiscal gain of extending the presumption to 24 months depends on how many additional cases this captures and whether probation capacity grows in parallel. On balance, the evidence points to a modest fiscal improvement — savings outweigh investment costs if the probation service is adequately resourced — but the magnitude is minor and the confidence is moderate given the delivery risk.

Crime, justice & national security — Helps

moderate · moderate confidence

Strong evidence shows short prison sentences lead to much higher reoffending than community alternatives, so shifting away from them should reduce crime over time. The main risk is that the probation service is already under severe strain, and an under-resourced system could undermine the gains.

The evidence

Biggest unknown: Whether the probation and community services will receive sufficient investment and staffing to handle increased caseloads, given current 21% vacancy rates and deteriorating offender outcomes.

Our reading: The evidence on the core mechanism is unusually consistent: short custodial sentences produce roughly 55% reoffending versus 32% for community orders (E2), and MoJ modelling projects over 30,000 fewer crimes annually from replacing sub-six-month sentences with community orders (E4). Restorative justice and diversion programmes add further crime-reduction evidence, with 27% fewer crimes among restorative conferencing participants (E6) and police diversion reducing youth offending (E7). The policy goes beyond aspiration — it commits to legislation (a presumption against under-two-year sentences) and investment in probation, meaning there is a real delivery mechanism, not just a soft verb. Short sentences represent 62% of all immediate custodial sentences (E9), so the population at stake is large enough for population-scale effect. The counterfactual (absent the policy, the current high-reoffending short-sentence regime continues) is well-supported. The direction is therefore 'improves'. Magnitude is 'moderate' rather than 'major' because the dominant risk is delivery: the probation service already has 21% officer vacancies, risk assessments adequate in only 28% of cases, and serious further offences by people on probation up 55% (E34, E37). If the promised investment fails to materialise or is insufficient, community supervision of a larger cohort by an under-resourced service could produce more serious further offences, partially or fully offsetting the reoffending gains. The time horizon is long-term: legislative change and probation capacity-building take years; reoffending rate shifts accumulate gradually. Confidence is moderate because the mechanism is well-evidenced but delivery risk is real and unresolved.