You don't need to jump down my throat with your obsession with data. You won't get data until someone decides it's worth trying, and the broken-ness of the welfare system is an pretty good basis to justify such an experiment. I'm sure if you dig up the data on welfare it can be used somehow to satisfy your statistical justification requirements, but keep in mind that public opinion is not as easily swayed by a good dataset as engineers are.
I'm not suggesting the necessity of small-scale experimentation solely for the purpose of convincing the public that this is a good idea, I'm also suggesting it for the purpose of convincing myself that this is a good idea.
As I said, I'm skeptical that this idea will work. It would however be fairly trivial (and cheap!) to carry out an experiment which could convince me otherwise.
Frankly, anyone who objects to the notion that we should spend $50 million or so to test the idea before implementing it on a large scale is a complete fool.
