Perhaps we could keep those reserves in some kind of locked box, so that it can't be abused ;)
Seriously, I think that expecting the bureaucrats to keep their hands off any funds left sitting around is way too optimistic.
Its a not-uncommon libertarian view that's probably more common the farther right you go (but you'll also find some left-libertarians who believe it; heck, you'll find left-libertarians that are almost completely opposed to the existence of the state as such)
"left", "right", and "libertarian" are all pretty broad groups within which there is lots of individual variation.
Fewer desperate people means you have the liberty to walk streets at night without the fear of being mugged. You are less likely to be panhandled by homeless people missing body parts because they couldn't afford their medical bills. Assuming BI suffices for the myriad of society's problems and replaces government solutions, it is an effective way to contain the state's scope and ambitions, since any "deserved" welfare scheme rests on the ideas that there are "correct" ways to live and "incorrect" ways that require welfare to fix. Basic income does away with that idea by providing what people need as the basis for life, liberty, and the pursuit of happiness, while not mandating a particular way to live (such as drug testing for food stamps or work/-searching requirements for the unemployed who would be shown the door if they preferred to live as an unpaid community volunteer). This is why BI is often attributed as a conservative idea.
You think that desires for nonessentials (or even luxuries) magically go away in such an environment?
Heck, a guy who was playing in the NBA a year ago was shot dead while pulling a home-invasion robbery this week. NBA minimum salary was $507K.
It is often modernly invoked in the exclusive context of when taxes or state programs harm people enough to put them into financial insecurity, but it applies to private finance as well, and is principally rooted in the hierarchy of needs, which can also be considered a growing degree of freedom - if you are stuck at the bottom of Maslow's hierarchy, you are prisoner to it.
http://www.libertarianism.org/columns/libertarian-case-basic...
However I don't doubt that some will accept it on a pragmatic basis, with the hope or expectation that it will lead to an overall reduction in violations of the NAP.
It might be better to directly subtract the benefits from any such program from the UBI. Then the only people who will even bother signing up for the other programs are the people whose total benefits from the existing programs exceed the UBI, which should be almost nobody. Then having demonstrated that fact in practice, those programs can be discontinued for lack of any real use and the savings can be used to provide slightly more UBI.
That's an interesting distinction, I think of myself as pretty liberal but maybe I'm more of a left-libertarian after all.
I generally support UBI (at least for now, I'll certainly re-evaluate my position as more studies like this one are done). And there are other traditionally "liberal" ideas that don't sit right with me, like supporting the old guard of e.g. taxi medallion owners, hotel owners, landlords & property owners (by opposing market-rate housing development), etc. even when the new alternative is actually better for the average lower/middle class citizen.
Actually, I think the two main groups are both libertarian, though there's a left-libertarian and a right-libertarian viewpoint, with the libertarians that aren't strongly left- or right- splitting between them.
I personally prefer keeping the existing means-tested programs but counting UBI as income when considering eligibility, such that with growing UBI [0] you eventually reach the point where the income floor is above the eligibility level for the means-tested programs, allowing them to be discontinued.
[0] I also prefer basically tying UBI to a tax stream that should grow with economic growth, with mechanisms to provide reserves so that you don't have UBI drop with cyclical recessions.