The "smart on crime" folks want us to believe that the experts know how to fix criminals, despite the disastrous consequences of our falling for that line in the 1960s.  This is not just in the United States.  Martin Bentham reports for the London Evening Standard:

A flagship Met police scheme to cut crime among convicts freed from jail has had no impact on the reoffending rate, an official study revealed today.

The £11 million "Diamond Initiative" was set up to rehabilitate serial offenders by offering them help with problems such as drug and alcohol misuse, housing, debt and unemployment after their release.

Scotland Yard chiefs hoped that the scheme, which focused on criminals freed after sentences of 12 months or less, would lead to a significant drop in reoffending and help to deliver the "rehabilitation revolution" wanted by Justice Secretary Ken Clarke.

An official analysis of the project has found, however, that 42.4 per cent of participants committed new crimes within a year of leaving jail - almost identical to the 41.6 per cent reoffending rate among a similar group of convicts who received no special help after being freed.

We have no objection to trying out programs on a limited scale and subjecting them to rigorous testing to see what, if anything, "works."  But until there is proof, not just "evidence," we should not abandon the tough sentencing practices that helped produce the dramatic drop in crime rates in the 1990s, and we should not return to the blind faith in the "experts" that produced disastrous results in the past.