How mandarins sabotaged a rare prison success story

Lara Brown Lara Brown
 Getty Images
issue 07 November 2026

Britain’s prisons are in crisis. So, naturally, Whitehall has managed to snuff out one of the few organisations that might know how to fix them.

Unlocked Graduates was founded in 2016 by Natasha Porter, a former teacher. Its goal was to solve one of the prison service’s most fundamental problems: the quality of its workforce. Prison officers have long been treated as glorified security guards. Applicants need not possess a single GCSE. Almost all candidates will never attend a face-to-face job interview and the governor does not meet those being recruited.

Unsurprisingly, this system has not yielded good results. Unlocked changed this. It persuaded ambitious graduates who might otherwise go into banking, consulting or the civil service to spend two years working in one of the most gruelling public service roles.

By 2024 the scheme had more than 40 applications per place. Around 55 per cent of successful applicants came from Russell Group universities. In contrast to ten weeks’ training that the average guard gets, those on the scheme undertook a funded master’s degree in leadership or custodial leadership.

But like all success stories born outside of the Blob, Whitehall soon grew suspicious. The Ministry of Justice wanted to turn the partnership into a conventional service-delivery arrangement. It still wanted Unlocked’s outcomes but also wanted to remove its ability to make that happen, essentially by moving the scheme in-house, binning the name and branding which was attracting talent, and prioritising a more conventional recruitment process. Contract negotiations fell apart and Unlocked closed earlier this year. The prison service finally had a way to attract the brightest into one of Britain’s least attractive professions. But Whitehall would rather things were done its way.

All hope for a better prison system is not lost. Unlocked leaves behind one big legacy: the Vanguard Wings. These are two specially managed wings inside two of the country’s most challenging prisons: Wandsworth and Pentonville. They were set up last year to evaluate the effectiveness of Unlocked’s model for recruiting, training and developing outstanding prison officers.

Porter says the approach on the wings is similar to a ‘strict but warm’ teacher. Prisoners are given clear boundaries and high behaviour expectations to ensure a standard of behaviour previously unseen in high-security, high-churn prisons.

In the past, prisoners have ruled wings. Guards were forced to work with the most violent gang leaders in the hope of placating inmates. The Vanguard Wings are different. If a remote control goes missing, for example, prisoners will band together to find it, because they actually respect the guards running things and know that bad behaviour is followed by real consequences.

Historically, Porter points out, our approach to rehabilitation is to assume that prisoners leave custody on the straight and narrow and then see if they reoffend. But the truth is that ‘many prisoners aren’t reoffending; they never stopped offending in the first place’.

If someone has spent their time on the inside racking up a list of infractions then, once released, they will almost certainly continue that same behaviour. The Vanguard Wings are unique because their high standards create an environment in which ‘men who have been habitually violent inside and outside of prison pause their offending’, sometimes for the first time in years.

Stats show the longer someone halts a pattern of offending, the less likely they are to relapse. The scheme’s interim findings also showed 45 per cent fewer assaults, a 57 per cent fall in self-harm and a 62 per cent reduction in disorder. Meanwhile, guards are taking roughly half as many sickness days and reporting a far better working experience.

Like so many areas of government, prisons have long been organised without reference to actual data on improving outcomes. Before the Vanguard Wings, a randomised control trial had never been used to test a prison-wing operating model.

There is, for the first time, a better way. But will Labour care?

Comments