I think there is a big problem. When you start a new program, everybody is very invested in it. They're excited about it. Maybe the participants are involved in it as well. They may feel they are part of a new venture, and this causes an effect that lifts the program. It used to be referred to as the Hawthorne effect after a factory experiment years ago, but it is a problem with pilot studies. You can do a pilot study and you may well get good results—a big drop in reconviction, improvements in jobs, people going back to school, people in work—so you've made changes beyond just the drop in crime. You have social changes. You have women going back to work and dealing with problems in the family, and the children now not having difficulty in school.
All of those are very valuable, but if you try to scale up a project, you may lose some of that excitement, and it becomes more routine, so this is a big problem, I think, for many projects.
What's happening in Britain now is that the model of the Peterborough, following people through prison and then afterwards, with five organizations working with them.... They're stopping the third year, in a sense, and financing it differently, because the government has a new plan for rehabilitation. Transforming reintegration it's called, so now everybody is going to get a year's support and follow-up, and they're sort of doing what the experiment was doing in a way.
Whether they will get the same results will be interesting to see, but they may not be as good. There is one academic called Aos, who has done a lot of work on cost-benefits and cost effectiveness in crime prevention, and he says you have to discount 25% of the effect of a pilot program, which is quite a lot.