Are one in 10 prisoners a former soldier?

Published: 16th Jul 2012

"One in 10 prisoners is a former soldier, new research reveals."

Independent on Sunday, 15 July 2012

The demands being placed upon soldiers was top of this weekend's news agenda after it emerged that many were being asked to fill gaps in the Olympic security team left by contractor G4S.

However the Independent on Sunday was among the papers to also run a story claiming that one in ten of prisoners were former service-personnel (also reported inthe ObserverSunday Times, and Sunday Express).

This figure is around three times the official government estimate of 3.4 per cent, so how can we explain the discrepancy?


Currently, official government statistics indicate that around 3.4 per cent of prisoners are former members of the UK armed forces.

This figure was arrived at after a collaborative research project published in 2010 by the Ministry of Justice and Ministry of Defence compared the records of those incarcerated in UK prisons with those who had left the military. 

The accuracy of this figure has since been challenged by campaign organisations such as No Offence which claims that the estimate fails to include many former military personnel who have been sent to prison after returning to civilian life.

For example, the MoD's dataset does not include soldiers who served in Northern Ireland, and therefore is likely to underestimate the number of former soldiers incarcerated in the UK.

Consequently, No Offence undertook its own research which concluded that around 10 per cent of prisoners are military veterans - the figure that made its way into the Independent's headline.

Unfortunately, neither the Independent nor the No Offence website explains the source of this figure beyond consulting "samples of the prison population". Therefore, we contacted the organisation's chief executive Sue Clifford to enquire after exactly where this number had come from.

Ms Clifford informed us that No Offence had surveyed a wide variety of research on the proportion of ex-servicemen and women in the prison population but had not produced a report itself that estimated a national figure.

In reality, it would appear that 'one in ten' is an informed guess based upon the evidence that No Offence were able to obtain. Ms Clifford made it quite clear that there was no known national figure (other than the Ministry of Justice's estimate) and that No Offence had heard multiple estimates from a variety of sources, generally ranging between 6 per cent and 14 per cent. 10 per cent marked the mid-point of these figures.

So where does that leave us?

The No Offence study isn't a detailed piece of quantitative analysis, so it is difficult to say with any certainty that it perfectly captures the proportion of British prisoners who are former members of the armed services.

That is not to say that the underlying message of No Offence is necessarily wrong. The charity has compiled a plethora of anecdotal evidence suggesting that the current estimate of 3.4 per cent is too low.

However whether the 3.4 per cent figure is flawed in its methodology is more moot. Ms Clifford suggests in the Independent that:

"The data captured when people go into prison is not accurate, as people do not talk about their former role unless asked. Some veterans, even when asked, do not say as there is huge pride associated with being in the military."

However the figure has actually been arrived at by comparing military records with prisoner data, taking the element of self-reporting out of the equation (although the problem may beset other studies on the subject).


Our investigation has left us unconvinced that it is actually a fact that one in ten of prisoners are former soldiers. Although there remains considerable uncertainty about the current estimate of 3.4 per cent, the 10 per cent proposed by No Offence (and consequently the Independent) seems largely to be an estimate based upon local and anecdotal studies, rather than a comprehensive look at the national picture. 

More research would be necessary before the 'one in ten' claim could be entirely substantiated.

Was this page useful to you? Yes  No


Full Fact and international partners win Google AI Impact Challenge

We aim for our factchecks to be as accurate and up-to-date as possible. If you think we've made an error or missed some relevant information, please email