It replaces more than just welfare. There are lots of different programs to prevent people from falling into poverty, whether due to long term unemployment, disability, retirement without savings, or part-time work at minimum wage. UBI can replace all of them. Progressive tax rates tend to have a 0% tax rate for the lowest bracket because people with a very low income need it all to just get by. If they get UBI on top of that, it could be an option to compensate that by raising the tax rate on the lowest tax bracket (and others maybe too).

The point is not simply to give people more money (though in some countries that might not be a bad idea), but to ensure that they don't lose that money if their situation changes. Often it's more attractive to remain completely unemployed than to work part time at minimum wage. With UBI, every bit of paid work you do, leaves you better off financially. It removed disincentives to work, because working people get it too. The costs of UBI are easily balanced with taxes. And because you don't have to qualify for it, you can save a lot on bureaucracy; it leaves people at the bottom of the economic ladder with a lot more freedom in their lives.

> UBI doesn't really make sense as a replacement to almost any traditional welfare programs for the simple reason that by limiting the money to only those who qualify for it, you can give more.

For means-tested programs, you can replace means-testing with unconditional benefits and appropriate progressive tax rate changes, and acheive exactly the same effects (modulo differences in definitions of income, but is there really a big benefit there?) while eliminating duplicate bureaucracy. Most UBI proponents don't prefer as rapid of a claw-back (which sometimes exceeds 100% in some ranges) as the aggregate of existing means-tested welfare programs, and would instead claw-back at a slower rate but extend the claw-back higher up the income distribution so that the drag is on the rich rather than the lower middle class, to improve upward mobility compared to existing means-tested welfare, but if you wanted to optimize only for bureaucratic efficiency without moving the claw-back higher up the income distribution and keeping it below a 100% marginal rate to eliminate perverse incentives, you could do that with UBI and tax changes, too, and still have a (smaller) net win.

The standard counterargument is that the current welfare system uses a shambling bureaucratic apparatus which forces people to go through an onerous, dehumanizing process of filling out endless forms to determine whether they're the "deserving poor" or the "undeserving poor". UBI, in dispensing with with this expensive process, is claiming back money that's essentially just being hurled into a void for no reason.

I have no idea whether that's enough to make up the difference.

Which is why ubi mostly make sense if there arent any jobs left not if there still is a substantial jobmarket.