Unsafe prisons cause crime

Research by Katherine Auty and Alison Liebling, just published (behind paywall – how can that be right, with research  done with public money? – but message me if interested) confirms that unsafe prisons don’t reduce reoffending. It takes measures for the quality of prison life (generated through anonymised surveys of prisoners’ and staff) developed years ago by Alison, and relates them to variations in reconvictions rates for those prisons, by excluding other variables. 
 
As a result, we can now say definitively, what common sense surely told us previously, that a safe, stable, decent prison is likely to reduce reoffending and a violent and chaotic one is not. It is the measures of safety and security that are most strongly associated with reduced reoffending, especially ‘prisoner adaption’ (the prisoners perceived need to trade or make alliances within the prison, presumably relating to drugs and so on): “for every one-unit increase in the prisons mean score for prisoner adaptation there would be a 10.67 decrease in rates of proven reoffending”. An astonishingly high figure, larger I recall than most offender programmes (of course, they are linked: if a prisoner feels safe, he/she will benefit more from such programmes). Other strong correlations were with measures of prisoner safety, policing and security, and drugs and exploitation.
 
The message for MoJ is that if prisons remain violent, chaotic places full of drugs, as is the case in many now, you won’t reduce reoffending. Indeed, I wonder if one could use this work to estimate how many more hundreds of thousands of crimes have been caused Grayling’s cuts? If anyone can manage this in a statistically convincing way, I am happy to raise funds to put up posters around his constituency!

Fig 1:  an intelligent Grayling