Like many things, there's substantial evidence on both sides of the argument. The devil is usually in the details, and this is a case where success or failure seems to be greatly affected by the quality of the programs themselves and the people who administer them.