One thing I worry about in terms of the social impact bond approach is something in social science that we call the Hawthorne effect, where the group that's being researched or investigated behaves in certain ways because of the fact that they're the focus of all this attention.
It's important to underline—and I haven't heard this today so far—that we do all kinds of research into effectiveness of service delivery now. We have randomized control trials and we have evidence-based research in social work and medicine in various fields. I guess one of the things I find missing from the social investment bond discourse is that we currently do have ways of measuring effectiveness of various kinds of services, and maybe the SIB approach is a bit of a distraction in some ways.