Research and analysis

Estimated progression model tranche release volumes

Published 30 August 2026

Ministry of Justice Analytical Report

This report sets out the methodology and estimates for the volume of offenders expected to be released on the first day of each progression model tranche release from October 2026 to June 2027.

Background

The Sentencing Act 2026 forms part of the Government’s response to the current prison capacity crisis, to avoid a breakdown of law and order by ensuring the expected shortfall in prison places against growing demand is addressed.

The ‘progression model’ is one of the Act’s measures and introduces a new release model for individuals serving Standard Determinate Sentences (SDS). Under the model, all offenders on standard determinate sentences will spend at least one-third of their sentence in custody, while those serving standard determinate sentences for more serious offences will serve at least half in prison. Those who behave badly can be held in custody for longer, with no upper limit on that additional custody time.

Following a review by the Government and alongside wider action to maintain sufficient prison capacity, the Government has made exclusions from the Progression Model. Rape, certain child sex offences and offences linked to organised child sexual exploitation – such as grooming gangs – as well as unlawful killing, and indecent assault will now be excluded. 

Other sentence types are also ineligible. Serious violent and sexual offenders serving Extended Determinate Sentences (EDS) will continue to serve at least two-thirds of their sentence in prison, with release determined by the Parole Board. Indeterminate sentences, including life sentences, also will remain out of scope. More than 18,000 prisoners serving these sentences are therefore not in scope for progression release.

The progression model is planned to commence from October 2026, and will apply both to eligible offenders already in prison and those entering the system in future.

To manage the release of the offenders from prison in a safe and orderly way, implementation for them will be phased across 10 ‘tranches’. Each release tranche brings the new law into force for a different cohort of offenders, based on sentence length.

On the first day of each tranche, an additional number of offenders will be released. This is because a set of offenders will already have served past their new release date when the law comes into force.

The estimated first-day release volumes of offenders in each tranche have been modelled to inform the implementation approach and are set out in this report. The Government has worked to avoid too large a group being released on any one day, so that releases can be planned and managed safely.

The Progression model is a new permanent approach. After each “first day”, releases for eligible prisoners will continue on a business-as-usual basis. For context, approximately 4,500 offenders are released from custody in a typical month.

The first-day release figures for each tranche in this report are modelled estimates rather than precise numbers. Future release volumes may change because: eligible offenders might be sentenced to prison before the relevant first day of releases for a tranche; some current prisoners have outstanding, additional criminal proceedings against them; and the number of added days for bad behaviour between now and the final tranche in June 2027 is not yet known.

Results

Table 1 shows the estimated number of prisoners who will be released on the first day of each of the ten progression model tranches, which commence from October 2026 to June 2027. It also shows the number of offenders who will no longer be eligible for release on the first day of each tranche, due to the introduced exclusions for rape, certain child sex offences, unlawful killing and indecent assault. As a comparison, the introduction of the previous change to SDS release points (SDS40) was split across two release tranches and resulted in 1,889 prisoners released on day one of the first tranche and 1,223 prisoners on day one of the second tranche.^{[footnote 1]}

Table 1: Estimated number of progression model releases on day one of each tranche and those now ineligible due to introduced exclusions

Tranche	Estimated releases on first day	Estimated number ineligible due to introduced offence exclusions
Tranche 1	700	50
Tranche 2	650	100
Tranche 3	650	150
Tranche 4	550	200
Tranche 5	600	200
Tranche 6	500	300
Tranche 7	200	150
Tranche 8	200	200
Tranche 9	250	100
Tranche 10	250	50
Total	4,500	1,400

The estimated release volumes in this table have been rounded to the nearest 50. Totals do not sum due to rounding. For context, more than 4,500 releases of determinate sentence prisoners take place in a typical month.

Methodology

These modelled estimates use different approaches for different cohorts of SDS offenders, reflecting differences in when sentence types were introduced, the offences they apply to, and their release points. Modelling cohorts separately allows us to account for these details.

For offenders who would previously have been automatically released at the 40% or 50% point of their sentence, the modelling uses a snapshot of administrative data for those in prison on 4 March 2026. This is used to identify offenders who would no longer be in custody had the progression model already been implemented, accounting for days added to sentences through adjudications.

For offenders currently released on Home Detention Curfew (HDC), the modelling uses an HDC caseload snapshot from the same date and estimates how the caseload would decay as offenders reach their automatic release date and move onto licence. This is used to infer whether offenders who would subsequently have been released on HDC (in a scenario where HDC remains in use) would already be beyond their progression release point, and therefore eligible for release in a progression model tranche.

For offenders serving an SDS with a current release point of two-thirds, the model uses a recent period of consistent prison inflows for this sentence type. By repeating these inflows and estimating outflow dates, it estimates how many offenders will be in prison and beyond the 50% progression release point during tranche releases for offenders of that sentence length.

The estimates are adjusted for the impact of other Sentencing Act measures, including the presumption to suspend and expansion of suspended sentence orders, and for projected changes in the prison population between the snapshot and relevant first day of releases for a given tranche.

As with any modelled estimates, there is uncertainty in the underlying assumptions and actual tranche release volumes may differ. These estimates are not equivalent to the prison place impact of the progression model – that is, the number of prison places saved.

Key assumptions

• Modelling categorises offenders based on their “main sentence”, identified by their most serious offence. If this does not identify the appropriate cohort, release volumes would be affected.
• Progression eligibility has been based on main offence only. Some prisoners may be serving secondary offences that make them ineligible, which has not been accounted for here and could mean modelled release volumes may be an overestimate.
• The modelling makes assumptions about factors that affect offenders’ release dates, including time served on remand and days added onto sentences. If these factors differ from the assumptions made, modelled release volumes would be affected.
• The modelling accounts for other measures within the Sentencing Act, such as the presumption to suspend and expanding suspended sentence orders. If their impacts differ from projected values, expected modelled release volumes would be affected.
• These estimates consider projections of the prison population between the snapshot date and the first-day releases of each tranche. As these are projections, they carry a level of uncertainty. If the actual prison population is different from that projected, then these estimates would be affected.

Key terms

• HDC: Home Detention Curfew allows offenders to be released from prison before their automatic release date, subject to strict curfew conditions and tagging.
• SDS: Standard Determinate Sentence – A type of immediate custodial sentence which means that offenders are automatically released after serving a set proportion of their sentence. They serve the remainder of the sentence on licence.
• EDS: Extended Determinate Sentence – A type of immediate custodial sentence for higher risk offenders which means that offenders are not automatically released until at least two-thirds of their sentence and require Parole Board approval for release before their sentence end date. Offenders are also given an extended licence period upon release.
• SSO: Suspended Sentence Order – A custodial sentence which is suspended meaning that the offender is not subject to immediate custody but can be sent to custody should they breach the terms of the order. The Sentencing Act expanded eligibility for SSO’s to those with sentences of 2 to 3 years in March 2026
• PTS: Presumption To Suspend – A measure implemented in March 2026, as a part of the Sentencing Act, to increase the use of suspended sentence for offenders sentenced to 12 months or less in prison.
• ESSO: Extension of Suspended Sentence Orders – A measure implemented in March 2026, as part of the Sentencing Act, to allow both the time a custodial sentence can be suspended for, and the length of custodial sentence that can be suspended to increase from two to three years.

1. Transparency data: SDS40 tranches 1 and 2 day one ↩