> However, programs which do not require meeting any qualifications are massively easier to implement, which can make them much cheaper overall -
I agree you can get some efficiency gains from eliminating thresholds.
Back of the napkin, though, you're talking about an implementation cost delta (ie, only threshold determination costs) that swamps the benefit-and all other implementation costs-by several multiples, maybe an order of magnitude, depending on the size of the targeted group.
But ok... Although it's not my intuition that threshold costs would significantly outstrip the savings from targeting, it's just a factual question. For the sake of argument, say that happens in a case.
Then yeah, that's a great candidate for universality.
If those are the cases you want to focus on... Are we then agreeing that universality is preferable only in those cases where it pays for itself?
Because that is totally consistent with my position. If you only want universality when it's cheaper then we already agree and can just defer to the CBO.