banner supportlinkscontactaboutarticlesmain

Prison Works

Please be aware, there are sections of this document that are still under development.

Here we report our reasoning for the claim that "prison works". Put simply, we believe reducing crime should be the primary concern when forming policy; everything else should be put second to this.

We base our assertion that "prison works" on a number of arguments:

* Statistical evidence from America, based on recorded crime, shows that as the liklihood of going to prison increases, crime decreases.
* Evidence from the UK, based on surveys and recorded crime statistics, also shows that increasing the liklihood of getting caught and being put in prison reduces crime.
* There is a clear economic case for putting people in prison, as it is not as expensive as the cost of crime.
* Community based rehabilitation schemes are no cheaper than prison, and are no more effective at reducing reoffending rates.
* However, prison based rehabilitation schemes significantly reduce the reoffending rates of repeat violent and sexual offenders. Unfortunately, these schemes are not the norm.
* Community based sentences cannot offer the caste-iron guarantee that prisons can - offenders cannot reoffend whilst in prison.
* We do not have a large prison population, when comparing them against the number of reported crimes.

Falling overall crime
Steven D. Levitt, Professor in Economics at The University of Chicago, finds that murder rates in America plunged 43 percent from the peak in 1991 to 2001, reaching the lowest levels in 35 years. The FBI violent and property crime indexes fell 34 and 29 percent, respectively, over that same period. Furthermore, a comparative study of the crime rates of US and England between 1981 and 1996, carried by Prof David Farrington of Cambridge University, showed that as the risk of being imprisoned rose in the US, the crime rate fell. Conversely, as the risk of being imprisoned fell in England and Wales during the same time period, the crime rate increased.

So what’s going on?

Two of the four main factors of the reduction in US crime are police numbers, accounting for up tp 20% of the decline in crime [note: it's similar for Europe]) and a rising prison population, with increasing prison numbers accounting for up to 27% of the overall reduction in crime. The total US prison population grew from 123 prisoners per 100,000 people in 1980, to 400 in 1996. In 1996, we imprisoned just 102 per 100,000 people.

Indeed, despite the rhetoric about short-sharp-shocks, during the 1980s the Tory Government pursued an overall anti-prison policy. The prison population was actually cut by nearly 10% between 1988 and 1993, even with a rocketing crime rate. The crime rate reached an historic peak soon after and towards the end of 1993 the policy was reversed by Home Secretary, Michael Howard. Within 2 years of reversing the anti-prison policy, Michael Howard witnessed a sharp reduction in crime rates. In his own words, “prison works”.

From the Home Office own statistics, if we plot the total number of crimes (British Crime Survey) as a function of time, there is a clear trend upwards to a peak in 1995, after which crime consistently reduces, such that in 2003/4 it is back to the 1981 level.

If we then plot the UK prison population, from National Statistics data, over the same time period we can see that the period of increasing crime corresponds to a relatively static prison polulation (it actually reduces by 10% between 1988 and 1993), and the turning point of crime is approximately 2 years after a tougher prison policy was adopted by Micheal Howard.

We have already commented that there are serious flaws in criminal statistics, as the BCS under-reports the more serious crimes. Unfortunately, there are also problems associated with recorded crime statistics - certain types of crime are woefully undereported. Perhaps a better way of presenting the data would be to plot individual classes of crime against incarceration rate - and this is something we are working on. Nonetheless, if we plot total report crime against time, the trends we see are remarkably similar.

Even now, with an expanding prison population courtesy of Mr. Blair’s administration, just 12 people are imprisoned for every 1,000 recorded crimes. In Spain, the number is 48. Spain's crime rate is roughly one quarter of ours. Indeed, if we perform statistical analysis of European crime figures, we find that those European countries which employ the sanction of prison the most tend to experience lower crime rates than countries which rely more heavily on community sentencing - nearly a third of the variation in European crime rates can be explained by the effects of prison.

Violent crime statistics
This will be added shortly.

One of the most prominent arguments against prisons is that reoffending rates are very high. In our attempt to find a solution to address this, we have done considerable research into community based schemes. Community sentences often take the form of rehabilitation - commonly referred to as ‘what works’ - and typically employed cognitive-behavioural approaches which seek to modify negative thought processes, which behaviourists hold to be at the root of crime. Programmes lasted for a fixed period of time, and operated on a clear medical treatment model, involving a fixed ‘dose’ of treatment leading to a ‘cure’ in a certain percentage of cases.

We have reviewed statistics on these schemes from academic and Home Office publications. We found official government reports with conclusions not matching the data contained within them and academic papers that displayed flawed methodology and inconclusive, incomplete statistics. We have found little objective or statistically significant evidence that community rehabilitation schemes reduce reoffending rates.

For example, if we take the first scientifically valid evaluation in England and Wales, of the 59 offenders sentenced to probation who completed the Straight Thinking on Probation Programme (STOP) with a control group of 100 offenders. Results from the 12-month follow up indicated that the STOP completers’ actual rate of reconviction was 8 percentage points lower than their expected rates (39 per cent as opposed to 47 per cent). However, the effect diminished after two years: the STOP completers had a rate of reconviction of 66 per cent and their actual rate was 68 per cent - statistically insignificant, given the sample size.

But one must also take into account the cost of sentences. In 2002 the Social Exclusion Unit estimated the cost of an ETS course of 40 hours to be £2,000 per prisoner and a course for a high risk, violent offender to be £6,000. At the time the average cost of a prison place for 12 months was £37,500. However, whilst the weekly costs of treatment were lower than those for custody, the greater length of time involved in community treatments often brought the total cost to the same level. Community interventions are not, in fact, a cheap alternative to custody.

Despite the worrying lack of rigerous evidence, and underestimates of cost, a major programme was introduced and in 2000/01- 6,000 offenders were put through accredited courses. In 2003/04 the Prison Service target was 8,900 and the Probation Service 30,000 in the community. Whilst it is most likely true that some community based schemes do reduce reoffending, we cannot find any evidence to support this conclusion - even if we contact award winning schemes directly. Furthermore, one must not forget that no community based scheme may offer the cast-iron guarantee that criminals will not reoffend whilst on the program, as prison does.

On the other hand, prison based rehabilitation schemes have been statistically shown to work - and is especially effective for particularly troublesome criminals - repeat violent and sexual offenders. HMP Grendon, near Aylesbury, is a specialist prison for males, run on the lines of a therapeutic community for those with a personality disorder. Prisoners selected for Grendon tended to be high risk offenders, when compared with other prisoners of similar age, serving similar sentence lengths for similar offences. Offenders spend long periods of time discussing, facing up to and accepting responsibility for the offence that led to conviction. They are also forced to examine and challenge the patterns of behavioural development that have led them to offend.

The research can be summarised by the following graph:

And for certain types of prisoner, a much larger effect was seen. For prisoners with two or more sexual or violent sexual convictions, the Grendon reoffending rate for sexual crime was 18%, with violent sexual crime being 31%. For the control group, the rates were 43% and 72% respectively.

The results of suggest that Grendon is not a suitable regime for younger, first-time violent offenders. However, it does seem to have an impact on the offending of groups often thought to be particularly troublesome, dangerous or beyond help such as older repeat sexual and violent offenders.

Caveats and myths
There tends to be the argument that the UK is a country with a high population of prisoners, yet we still have a relatively high crime rate. Despite this misleading rhetoric produced by the anti-prison lobby, our prison population is not large in comparison to other European countries. Yes - we have a large absolute prison population - but this is small compared to the number of crimes. Comparing absolute numbers of prisoners is a very misleading way of representing the statistics.

One must also be careful of any claims that prison does not work, when this is only backed up with a 'length of sentence' or 'likelihood of going to prison' argument. For example, over a 10-year period, the percentage of convicted burglars sent to prison rose from 37% to 60%, and for grievous bodily harm the figures are 28% rising to 54%. When presented by statistics such as these, the argument goes "But crime rates have not fallen, so prison can't work". However, if the likelihood of getting caught plummits, this over-simplistic argument is nothing more than a gross misrepresentation.

One of the favourite arguments against increasing the prison population is cost. Prison is extremely expensive - during 2003-2004, it cost an average of £27,320 per year to keep someone in prison and to build a new prison, it costs the equivalent of 2 district hospitals or 60 primary schools. But the cost of crime which is frequently ignored when citing these numbers. In 2000, The Home Office estimated the cost of crime at £60bn. Taking inflation into account, it's £3,000 pa for every household. Doubling the number of prison places would cost £7bn - so it would be revenue neutral if doubling the number of prison places reduced crime by a little over 10%. If one takes a purely economic perspective, prisons are not as expensive as crime.

In conclusion
There is a correlation between countries with a high likelihood of being caught and incarcerated, and low crime rates. This clear relationship is often denied by those that think prison does not work. Simply put, it is the likelihood of being caught and put in prison that related to the crime in a country.

It is clear: to reduce crime, the liklihood of getting caught and being incarcerated needs to be high. In this particular situation, the sum is greater than the parts.


How do I get in contact with the people behind this site. I would love to use you as part of a radio interview this evening.

Dear Peter

I have sent you an email in response to your query. I apologise for the lack of details on the site - but this site has not been officially launched yet. I expect we will make a press release soon.

Thank you for your interest and I look forward to corresponding with you.

Post a comment