[dllthomas]

It's none of the cases dragonwriter described, but what seems to be closest to what you're analyzing:

If you have a fixed UBI, paid for with a variable tax that splits the total burden of the UBI across all workers in proportion to their productivity, then you would see dynamics like you describe. This is, indeed, the opposite of a negative feedback loop. Don't do this, either.

If UBI is set high enough that more people drop out of the labor force than we can manage, the (real) UBI must fall by some mechanism.