They really, really don't. All the experiments on a small scale show is the trivial conclusion that if you give people some central government funds for a year or two, they tend to be slightly less poor at the end. None of them come remotely close to suggesting that people are so much more productive that they can cover the costs of UBI to people already out of the labour force by choice, never mind those that may choose to exit it in future.
Elderly people already get a UBI in the form of Social Security, and that's politically impossible to cut regardless of whether it's a good or bad idea.
Once you turn it on and people get used to it, you can never turn it off in a democracy.
Only if the country still has the ability to change and the economy isn't permanently damaged.
There's this set of objections to UBI based around "yes, ok, it works in the small scale, but it'll never work on a large scale".
If we did a large scale experiment, and it confirms that it works, the objections will move to "sure, OK, but it'll never work on a national level".
So let's just implement it and see if it works. If it doesn't work, we can always just go back to letting people starve.