Those are good questions, but I apologize. For some reason, the translation didn't come forward on the first component of your question. You have my apologies, but I'll answer it to the best of my ability.
Your observation is astute, that at this point in time there is not a great deal of results or data or outcomes to draw full conclusions on, in any area of social financing. In the case of microloans, I would say, absolutely, that those have been around for decades and the outcomes are proven. For social impact bonds, though, your point is quite a good one.
I say that in the absence of perfect information, and these are some of the limitations you were speaking of. I think we have a responsibility to look at a situation and, on balance, with the information we have and with what we know, ask whether it is plausible, whether it is possible, whether we are convinced that we are going to see an outcome that is as positive from a financial perspective as it is from a social perspective, whether it could be achieved in another way, and whether there's a positive business case for it. If you have that positive business case and it's compelling to government, that is evidence and reason on which to rationally decide that perhaps this is something we should try. We should try to ameliorate, in certain groups, the graduation rate or the recidivism rate or the rate of retention of at-risk individuals.
As well, we have a responsibility to objectively study it during and after, so that if the results are not positive, we learn from that and adjust programs, or we learn from that and decide that maybe this isn't a good mechanism for certain services or certain sectors. I think the responsibility is clear for the predictive business case that we should proceed only if it is compelling and strongly positive, and that there should be measurement and follow-up and transparent reporting back to taxpayers about the results.