[pwinnski]

Any test that doesn't involve full nationwide UBI is subject to the same criticism, and I predict that even a large-scale three-year-long test that involved every western state will still be dismissed as unworkable for the same reasons: no barrier between eastern and western states, and everyone knew the test would end in three years.

[luckylion]

> Any test that doesn't involve full nationwide UBI is subject to the same criticism

No, not really, unless the money for the experiment comes from outside the region. The problem is that the argument for UBI is either that "it'll pay for itself" or "it doesn't cost more than today's system". So we'll need a test that actually tests that.

Maybe it doesn't work and we'll see "okay, it'll be 10% more expensive than today's systems". That can be achieved by raising taxes, and the result of doing that can be seen in the experiment as well: do people leave the region because they don't want to pay an extra 10%, or do they find UBI's benefits to society at large to be so large that they don't mind?

Maybe we'll see that UBI does what the proponents promise: free human innovation and productivity and easily more than pay for itself. Once that result is proven, you'll have no issue to convince anyone.

Not trying to test that at all and then saying "well, even if we would, nobody would accept that" isn't the right approach in my opinion. And it feels like there's something left unsaid: that the proponents also don't believe that it would work and therefore don't want to actually test it, because they're convinced the test will prove that it doesn't work. And as long as it's not tested, they can claim that it totally would work (in theory) without having to prove it.

[pwinnski]

So how do you propose a better test? It seems like you slid easily past that point. To determine whether or not it requires 10% higher taxes, as you suggest, we would have to conduct a test large enough to have an appreciable impact on tax allocation, which seems like it would need to be a really wide test, like state-wide, that lasts for a at least a few years, no?

I guess that's what I'm not seeing: any tests done so far are too small and not long enough and are known ahead of time to be a test. Okay, fine, let's conduct a test that's larger and long-term, with an eye toward permanence. How large, though?How do we avoid the criticism seen on this very page that limiting the number of people involved distorts things? How do we avoid the converse, which is that unless you're Alaska, letting it be known that everyone in a certain area gets free money means an artificial boom in that area?

[luckylion]

I don't think it's fair to ask those that are skeptical of a claim to provide a test that proves it doesn't work, the duty to prove is on those making the claim.

But still: yes, you'll need a larger unit. It won't necessarily have to be a state, though that depends on the country and the tax-setup within. It would have to be something that can set their own taxes. A village would likely be too small, but a medium sized city of 20-50k should certainly see the benefits if they exist. And it wouldn't need to be limited to a few years: convince the inhabitants and you can democratically vote it in, it'll run in perpetuity, or until the money's gone in the case that it doesn't work.

> How do we avoid the criticism seen on this very page that limiting the number of people involved distorts things?

That's not really what's criticized. The issue lots of people have with these tests is that they're only testing the distribution of money, but not the funding. You don't need to have millions of people. I'm pretty sure if you can get 10000 people that are broadly representative of the population at large and get them to fund their own UBI, collectively, and play it out, that'll be a good test in most people's eyes. I'll certainly pay attention, because it'll actually test UBI, not just "if we take money from the national taxes and give it to 500 people, what will happen?"

> How do we avoid the converse, which is that unless you're Alaska, letting it be known that everyone in a certain area gets free money means an artificial boom in that area?

An artificial boom would still be a boom. Would people invest in an area if they knew that they'd have to pay high taxes? If so: great, let them do it, that's not so bad. The problem arises when the number of people asking for UBI grows faster than those investing & funding the UBI. But that's exactly what has to be tested, because it's pretty obvious that the same would happen on a state or national level - unless we're talking about closed borders, which sounds anachronistic.

[pwinnski]

Thanks for the detailed reply!

> It would have to be something that can set their own taxes

In the US, at least, income taxes are levied at the state or federal level, not anything more local than that. Even sales taxes are generally limited by the state, if not set outright. I haven't lived in every state, so it's possible that there are exceptions somewhere, but generally I think such a test would have to come from elsewhere.

TL;DR: the smallest unit in the US that can set their own taxes is a state, making any test smaller than a state somewhat unhelpful.

> I'm pretty sure if you can get 10000 people that are broadly representative of the population at large and get them to fund their own UBI, collectively, and play it out, that'll be a good test in most people's eyes.

Ah, this strikes at the game theory heart of it! And yet still seems to be unworkable in the US. A representative cross-section would need to include both rich and poor, as the nation as a whole does. Either all are asked to pay additional taxes over and above their normal taxes, or the state in which they reside is asked to do with less tax income than normal, as some or all the taxes of those 10k would instead be redirected to the UBI trial. Both seem like a tough sell, either to the richest of the cross-section or to the state itself.

> unless we're talking about closed borders, which sounds anachronistic

Speaking of the US, while we're very poorly suited to doing a small-scale test, we're actually reasonably well suited for having somewhat closed borders, on account of being separated from most countries by big oceans. Movement within the US is trivial, while movement into or out of the US is very difficult. Presumably UBI would have some measure of buy-in that would exclude people who cross the border without documentation, at least initially.

It's even possible that we would end up with a situation like we have now, in which people working without real documentation pay taxes into social security, while never having the ability to draw on social security later in life.

Given the difficulty of finding a cross-section of volunteers and a willing state government, I wonder about alternatives. I think a case could be made that funding a UBI trial from normal revenue could still be instructive IF people participating in such a trial were opted out of all other benefits. That is, let's test the claim that UBI would be cost-effective in part because it would replace existing programs. Pick a cross-section of people, track how much is sent to them and how much is received from them, and make it closed system by denying them access to other support programs. We should be able to track what difference that makes for revenue and expenses, no?