The way the bonds work is that, if they are successful, the government pays a premium. It pays a return anywhere between 5% and 30%. So in terms of ordinary costs of borrowing, clearly this would be higher than that. The rationale is that the government would be saving much more than that. My argument is that this should apply to most government programming anyway. I believe that this does add an extra cost to government, and it's one that could and should be avoided, if possible.
On March 26th, 2015. See this statement in context.