“Britain risks descending into what the 17th-century philosopher Thomas Hobbes described as a state of nature, with an enfeebled state unable to prevent catastrophic breaches in civil peace”. John Gray, October 2025
In his New Statesman essay, John Gray describes - without descending into the civil war phantasma of the online right - the plausible consequences of a decline in state capacity, and the collapse of trust in state institutions in a society increasingly alienated from itself.
For all the welcome optimism of Andy Burnham’s first two weeks, the foundations of the state he inherits are hollowed out. His predecessor, Keir Starmer, entered power with an explicit promise to fix those foundations and that the government would do more to end the crisis of insecurity. Not enough was achieved. Whilst devolution and the economy have been Burnham’s central focus, the crisis of insecurity has not been resolved. Disorder is likely to return to Britain’s streets. Small boats continue to flow, and whilst asylum hotels are closing, HMOs stir up local anger. Whilst crime is decreasing, popular perceptions of crime rates remain understandably high, driven by a visible increase in certain crimes, and the degradation of the public sphere. The police, demoralised by scandal after scandal, are barely present.
Many of the institutions that uphold our security are broken but the remedies selected to reinforce them often sap their strength and undermine them. This series of notes attempts to engage with those problems head-on. We start with the prisons crisis, rebutting the core contention that Britain is ‘addicted to locking people up’, and arguing for a more strategic approach to early release. We then propose a number of short-term measures that may help to alleviate the crisis. A list of our recommendations is annexed at the end of this article.
To discuss any of these in further detail, please get in touch labourrealists@proton.me.
Notes on security: improving our prison system
Just as with Indefinite Leave to Remain, there are no attractive options to Britain’s prison crisis. That is the inheritance of 14 years of Conservative neglect and maltreatment of the basic institutions of Britain’s security. Just 500 prison places were added to our prison capacity in that time, despite necessary increases in the length of sentences for serious offenders. The new Justice Secretary Alex Norris needs to make that point repeatedly and assertively as he deals with the crisis left behind.
It is right that Burnham paused the early release programme ahead of 2 September’s first release. He should now announce the delay of the scheme in full so that its scope can be significantly narrowed. The core reality is that, against a backdrop of failing public protection services, the large-scale release of high-risk offenders is likely to lead to widespread re-offending and the re-traumatisation of victims, damaging dwindling confidence in the criminal justice system.
Whereas the previous early release scheme (Standard Determinate Scheme 40%, or SDS40) released prisoners at the 40% mark, it excluded serious offenders such as perpetrators of serious or domestic violence and sexual offences. The new scheme, which releases offenders at the one-third mark, has no such provision. The previous response to criticism of this was that it excluded the 18,000 prisoners who received an extended determinate sentence or life sentence. If a judge assesses that the test for dangerousness is met, an offender receives an extended determinate sentence. That is not, however, a foolproof way of determining risk to the public. It would not be accurate to state that there are not serious public protection concerns about the release of many of the remaining 60,000 prisoners. The large-scale blanket application of a scheme like this will mean that high-profile offenders are released early, with an extremely strong chance of a high-profile crime committed by one of them. The inquest and public apology are certain to follow.
The scale of crisis in probation and community management of offenders cannot be overstated. One officer who deals primarily with Registered Sexual Offenders describes offenders being found in bail hostels with devices in breach of their conditions, having brought them from prison. There has been an 8% increase in RSOs in his area, and each officer is dealing with an average of 64 RSOs, above the target of 50. Their shifts regularly exceed 16 hours because once an RSO is arrested, there is no one else to process, investigate, and charge that individual and prevent him returning into the community. This drastic anecdotal picture is supported by the wider evidence base on community management. In 2022-2023, HM Inspectorate of Probation found that for those classified as medium risk, 74% of assessments were not to the required standard.
As Ian Acheson has noted, our assessments of risk are built off an incomplete picture and are “dangerously superficial”. The majority of murders committed by those on probation are committed by those assessed as medium risk (68% were committed by those assessed as medium or low-risk), including the killing and sexual assault of Zahra Aleena by Jordan McSweeney, and the killing of a woman and three children (and the rape of one of them) by Damien Bendall. McSweeney had been released on licence from a prison sentence for criminal damage, racially aggravated harassment, and possession of a knife just nine days earlier. Bendall had committed eight offences, including serious racially aggravated violence, and had repeatedly breached the conditions of his licence. Wrongful classifications can and do happen regularly, with serious consequences.1
Contrary to protests, Burnham had no chance but to suspend the scheme which would have given piled pressure on an already failing system. It is not giving in to media hysteria or right-wing populism to acknowledge that this approach was not sustainable. It is true that opposition to ‘soft’ sentencing whether it be community sentences or probation is driven by harsh public attitudes towards criminals but this cannot be disentangled from public concern about horrifying protection failures like those set out above.
Whilst the most prominent critics of the current approach are right about the squalid nature of our prisons, and the unacceptably high reoffending rate of those on short-term sentences, they are often not engaging with reality about who is in our prisons today. David Allen Green recently suggested a supposedly countervailing view: “imprisonment for reasons other than public protection or for the most serious of crimes is not a public good.” Alan Rusbridger has suggested that “we’re addicted to locking people up”. The evidence does not substantiate this.2 Rather, the prison population can be divided into two cohorts: those who have committed serious offences, particularly violence against the person including sexual violence, and those who are prolific low-level offenders, serving short-term sentences because the system has very few other available options. That Britain has a high number of prisoners per 100,000 of the population compared to our European neighbours means relatively little as an isolated statistic. They too, including Sweden, are experiencing similar struggles with prison capacity, and sentencing. The fatal flaw of the Sentencing Review is that it places reducing the prison population as a primary goal of criminal justice strategy. As Tony Blair once stated, “what you cannot do is say that you reduce the prison population as a matter of principle irrespective of what criminal offences are being committed.”
What is to be done?
Below, we set out a series of measures that Norris should adopt to ease the scale of the crisis in the short-term. Before we turn to this, it is important to stress that the expansion of our prison capacity lies at the heart of any solution to the crisis in prisons. The lack of gainful rehabilitation, the proliferation of drugs and weapons, the squalor and violence, all of which strips dignity from prisoners and prevents them from making progress towards reintegration - these are all a direct consequence of prison overcrowding. The government aims to build 14,600 places by 2031. However, as David Spencer has noted, many thousands more are needed to address this crisis and ensure that prison capacity is not a determining factor for sentencing or policing approaches. The long-term ambition must be the large-scale increase of our detention capacity that eases the scale of overcrowding, and bolsters rehabilitation within prisons.
Limiting early release - who to exclude?
In light of the scale of the crisis in probation, it is right to exclude as many high-risk offenders as possible. Reporting indicates child rapists and other perpetrators of serious sexual abuse are likely to be exempt from the early release scheme, particularly bearing in mind the aforementioned crisis in monitoring RSOs. This would be the right approach for public protection reasons. Yet, regrettably, decisions on release cannot simply be matters of public protection. A level of unquantifiable risk is present every time an offender is released from prison, whether that be at the one-third, halfway, or at the very end of their sentence.
On the question of who else should be excluded, thought then should turn to the political mission and narrative. What is Labour trying to achieve in criminal justice? It has committed to reducing violence against women and girls by half, an ambitious and at present unrealistic target. That gap between ambition and reality can be closed further by restricting those convicted of domestic abuse from the scheme. Those who have harmed emergency service workers should be ineligible for early release, a small but symbolic gesture to the police that the government is on their side.
Two further measures are necessary. First, prison governors should be given the authority to block early release for specific prisoners on public safety grounds. This would require careful legislating, but, unlike many of the other measures we recommend that are rooted in solving a short-term crisis, this should be part of a long-term ambition of restoring the authority of individual prison governors. The broader principle at play is an important one: switching control out of Whitehall and into local decision makers. With governors having to balance extremely limited space and public protection concerns, it is unlikely the power will be used on a spurious basis.
Second, the government should review where sentencing has been expanded in recent years beyond the original intentions of Parliament. Following a terrorist attack in Fishmonger Hall in 2019, and a subsequent failed attack in Streatham in February 2020, both committed by recently released prisoners, the Conservative government introduced emergency legislation in the Restriction of Early Release Act 2020, imposing mandatory two-thirds prison sentences on all terrorist offenders. This was necessary in relation to Islamist offenders because the best means of preventing a suicide attack is incapacitation, and the risk posed by Islamist offenders is significant - likely to lead to a loss of life. Moreover, the longer a sentence, the more likely it is that an offender desists from terrorism.
However, this law is now beginning to impact a different cohort of terrorist offenders, those who conduct property damage and low-level violence for political purposes. Their activity does not come close to posing the same level of risk to public safety. We are applying a law intended to prevent suicide attacks to those whose offending can be managed outside of prison, freeing up prison capacity. Whilst this cohort is presently small, it is likely to significantly expand in the coming months, impacting tens of prison spaces. The government should explore a legislative change to apply the two-thirds rule on a case-by-case basis with terrorism offences, with decisions made by the Joint Extremism Unit which is well placed to make assessments on this.
Reducing recalls and partially reversing the Fixed Term Recall 56 day period
Between January and March 2026, 12,977 people were released from prison, but 13,193 were returned for breaching their licence conditions, an increase of 31%. For every 100 people released from prison, 101 people are recalled to custody.
The Sentencing Act 2026 ended the use of 14 or 28 day recalls, and set a fixed term of 56 days.3 The purpose behind this was to ensure there was sufficient probation time to risk assess offenders and understand why they had been returned. This followed a recommendation within the Independent Sentencing Review. However, the Review also recommended tightening the threshold for recall so that it was used to address consistent non-compliance with licence conditions, or specific and imminent risk. This did not happen.
Recall now risks being the predominant driver of a lack of prison capacity. Whilst this is evidently necessary for the 22% of those recalled for further offences, and others who do not commit crimes but breach conditions in a concerning way, it is counterproductive to impose the 56 day recall penalty on every individual who falls foul of their licence conditions. Those who are recalled and then released start from square one, with any previous progress on housing and employment wiped out. There is little merit to having a blanket 56 days to risk assess offenders when the cause of recall is sometimes likely to be known and unfixable, often caused by various gaps in service provision outside of prison that will not be remedied.
We recommend exploring two approaches. First, the Sentencing Review’s recommendation for tightening the threshold for recall should be adopted in relation to low risk offenders. Absent other risk factors, one-off mistakes or errors of judgment should not trigger prison recall. It is important to account for the fact that the systems designed to support offenders once they leave prison simply do not exist at the requisite scale and teething problems are likely. Probation officers should be empowered and encouraged to use their own discretion. Second, we recommend the introduction of a short-term recall that removes an individual’s liberty for a period (14 days) but is designed to avoid the longer-term 56 day penalty, reducing the disruptive impact on their life and freeing up prison space.
Reducing the flows into prison
It is accepted that prison is a justified option in some cases of non-violent offending such as financial crimes. The two hackers who caused tens of millions of pounds of damage by hacking Transport for London were each, rightly, sentenced to five and a half years in prison each. Yet amidst a context in which violent offenders risk not being detained, it is worth exploring measures to prevent prisoners such as Flowers and Jubair taking up prison space. In Norway, a queuing system is used to manage limited prison capacity. If imposed solely on non-violent offenders with concurrent substantial monitoring conditions, similar to the recommendation made by JUSTICE, this could be a method of ensuring prison capacity is available for those who pose a greater risk to public safety. Any breach of conditions would result in the offender being remanded to custody.
Expanding capacity in the short-term
On entering government Labour scrapped the Conservatives’ plan to rent prison space abroad. This was a mistake. With HMP Dartmoor (roughly 650 spaces) closed until further notice due to high levels of radon, and major prison building projects not due to complete until 2031, the need for short-term gap filling in capacity is clear. Labour’s pitch is that it is building prisons at the fastest rate since Victorian times and that until then, drastic measures are needed. If we want to keep voters on side, those drastic measures cannot solely be releasing prisoners prematurely
Whilst prison reform organisations are opposed to this idea, the majority of their concerns are unfounded. The Howard League for example has claimed that Norway abandoned their renting scheme after only a few years, ignoring the reality that it was only ever a temporary solution to domestic overcrowding issues. Another concern relates to the fact that only a relatively small number of cells are available abroad, or that it would be expensive. But, in the context of the release of 220 child rapists, 344 adult rapists, and 982 perpetrators of sexual assault on children, even 600 spaces would go some way to reducing the risk posed to the wider public. Re-offending too, has a vast and often uncaptured financial impact on society.
Any remaining concerns relate to the rights of individual prisoners such as the risk of an accountability gap, or difficulties in prisoner visitation. These are all legitimate concerns but they come second to the rights of the public at a time of acute crisis in our prison system. To allay concerns among the Parliamentary party and prison reformers, the enabling legislation should have a five year sunset clause, taking us to the point at which our prison capacity should have expanded. Foreign national offenders who are set to be deported at the end of their prison sentence should be prioritised, reducing (but not eliminating) the harm caused by imprisonment abroad.
Conclusion - Managing the crisis
Public confidence in criminal justice is fragile at the best of times. There will always be failures of protection - that is the nature of any uncontrollable risk. Too little thought was given however to how this scheme would be communicated to the public. The former Secretary of State for Justice did not appear to be interested in the implementation of his brief, occupied instead by his Deputy Prime Minister role. Whereas his predecessor, current Home Secretary Shabana Mahmood, had the political foresight to pair unpopular releases with popular tough-on-crime measures, announcing them quite boldly on the same day, there has been little attempt to sell this second wave, despite it being far more expansive. Mahmood also emphasised the time-limited nature of this approach, committing to reviewing it in 18 months time when the situation was likely to have stabilised.
There is little time to lay the groundwork for an identical approach. But the Ministry of Justice can be bolder. It is not enough to simply provide reactive lines that can be quoted at the end of a hostile Telegraph article. The lesson from Mahmood’s time at the department, and subsequently at the Home Office - which Norris played an important part in - is to shape public opinion through policy announcements which show that you understand and connect with what Philip Gould described as the populist instinct of voters.
Presently, the Ministry of Justice wheels out its core message: building new prison places; launching an expansion of tagging; and giving victims a say in licence conditions. None of these are objectionable, but they are unlikely to supersede headlines about rapists being released onto the street. They also lack the level of detail that justifies emergency measures in the short-term. When are prisons coming online? What emergency measures are we taking to get them built? Insurgent government is about setting the agenda and the tempo of your brief.
Norris must go on the offensive. Policies such as overseas prison rental might be expensive, and naysaid by Ministry of Justice civil servants, but they serve to both address the crisis and communicate its scale and the seriousness with which the government is responding. An expansion of the use of tasers in prison, higher standards for prison guards to end the spate of misconduct and corruption - these policies all have costs but build public confidence that their concerns are being listened to as well as meeting a genuine need to improve the security and quality of prison guards.
It would be a mistake to be complacent about the current nadir of Labour’s political opponents. Their return will come out of the next public protection scandal, or the next inquiry, fought on the battleground of their traditional issues such as crime or security. This need not be uncomfortable ground. Labour have fought and matched their opponents before on criminal justice, when they were prepared to approach the issue with requisite realism. But the government must be proactive and decisive by taking control of this policy issue now, and repairing public confidence in the criminal justice system the Conservatives broke.
Annex A: Policy Recommendations
Make expanding prison capacity beyond the Government’s own target of 14,000 a long-term ambition.
Exclude sexual offenders, those who have committed domestic violence, and those who have injured or killed police officers from the early release scheme.
Empower prison governors and protect the public by giving them the power to exclude specific prisoners from the early release scheme.
Change the Prevention of Early Release Act 2020 to permit non-violent, low-level terrorist offenders who do not pose a serious risk to the public to be released in line with other offenders.
Reduce the recall rate by encouraging parole officers not to trigger recalls for single breaches of licence conditions, absent other risk factors.
Reintroduce a shorter period of recall (14 days) to prevent recall reigniting the prison capacity crisis.
Introduce a queue system for those sentenced to custodial sentences but whose offending is non-violent and is manageable outside of prison.
Re-introduce legislation abandoned on entering government that would have permitted overseas renting of prison space, with a sunset-clause for five years.
Turn the tables on the debate by introducing new policies such as taser rollout, and higher standards for prison officers.
There is some limited evidence to suggest probation has become much more risk averse as a result of some of these failures, contributing to the high recall rate.
See also David Spencer’s Policy Exchange report that details the multiple offences needed to secure convictions against low-level offenders - https://policyexchange.org.uk/wp-content/uploads/Build-baby-build-A-new-generation-of-prisons.pdf
Its provisions came into play on 31 March, so did not impact the previous recall figures above.