It was in 1955 that it began to be safer to be a criminal in Britain. As Charles Murray, the American sociologist, pointed out, it was the year when the ratio of prisoners to crimes began to drop slightly. So began a sustained decline in the use of imprisonment that continued for almost 40 years. During that period the number of reported crimes increased 12-fold, while the number of prisoners merely doubled. Murray called this phenomenon “the great decline”. In the United States, much the same trends with higher crime rates eventually produced a fierce backlash: the authorities started to lock up offenders in ever greater numbers. They reasoned that if persistent wrongdoers were caught, convicted and locked up, then the public would be safer. Sure enough, crime began to fall. This was noted on our side of the pond and by the 1990s one brave Tory home secretary, Michael Howard, picked up the baton. “Prison works,” he famously announced.

Yet the human spirit quails before this brutal logic. It means expensively incarcerating tens of thousands of (mainly) young men and releasing them only when age has exhausted their appetite for crime. It means admitting that most schemes to rehabilitate offenders do not work. Men like Mr Murray argue that the solution is a lot less expensive and destructive than leaving them free to rob and kill. But many, knowing that the old liberalism has failed, would still like a third way out. Writing off a generation to prison is a bleak prospect. The electorate responded in 1997 to the message that you can be “tough on crime and tough on the causes of crime”. But can you really nip criminals in the bud? New Labour was elected to office pledged to try.

Some take hope from Britain’s muscular economy on the principle that the devil will find mischief for idle hands. Others look to reforms in welfare which get young men to hold down a job and not a dole ticket. Police procedures can be revolutionised. New York famously put its cops back on the beat and operated on the broken window thesis: if petty crimes went unpunished then worse would follow. In Britain, the trends which saw police taken off the beat and put behind the wheel of a fast car have begun to be questioned. Zero tolerance is still in its infancy here. And then there are the schemes to save children from taking a wrong turning.

Today we report that the Home Office is to send 70,000 potential young offenders on “positive activities” during the school holidays. Those who have been referred by police, juvenile courts and education authorities will be packed off on trips out of town. Ministers claim they are not bribing young offenders to behave, citing the success of pilot programmes in bringing down street crime. But the government has yet to prove that its scheme works. It has already tried several wheezes to reduce reconviction rates. They have failed. As Civitas, the think tank, reports, neither the Home Office’s intensive supervision and surveillance programme nor the electronic “tagging” of offenders have produced better results. Expensive “cognitive skills” courses have been introduced based on the idea that criminals carry out crimes because of mistaken beliefs. The Home Office now quietly admits that they are useless; 75% of Britain’s young offenders reoffend within two years of release.

Oliver Letwin, the shadow home secretary, argues that the seeds of future offending are sown in infancy. By the age of five, 15% of children display early signs of behavioural problems and are rejected by their parents. Half of those will go on to be criminals, while of those who become repeat offenders, more than 90% showed severe antisocial behaviour in childhood. The evidence indicates that bad parenting is the most important cause of a child turning bad. But the breakdown of the family is not going to be reversed by this government or the next. Politicians cannot bear too much reality. The next best thing it can do is to identify as early as possible when children are going wrong and then act. Perhaps this is the art of the possible.