The AI Threat Isn’t Skynet – It’s the End of the Middle Class

Because that's what a working safety net does. It prevents employers from imposing unreasonable terms on employees when their only alternative is starvation, by making that not the thing that happens if they turn down the job.

> There are many ways of designing welfare systems to avoid the "welfare trap" of excessive effective marginal rates of income taxation, of which UBI is probably the least efficient.

I'm going to describe two systems.

In one there is a UBI of $12,000/year and a flat 30% tax rate. In another there is $12,000/year in cash social assistance with a 20% phase out up to $60,000, a 10% tax up to $60,000 and a 30% tax over $60,000.

These two systems are in fact the same, the only difference is that in the second one when the government takes 20% of each additional dollar you earn they call it "phase out" instead of "tax", which leads people to the mistaken impression that increasing that rate (which applies to lower income people) is a sensible thing to do, even though it is identical to raising taxes on lower income people.

The second system is not "more efficient", it is exactly the same. And anything that doesn't look like that is going to be less efficient.

You can give people non-cash, and then you need an inefficient bureaucracy dedicated to busting people who figure out some way to divert their food assistance money into paying their insurance premiums or similar, and at the same time you create an inefficient barter system where low income people subvert that bureaucracy by converting the things that can be bought with government money back into real cash.

You can try to phase out the UBI, but as above that is completely identical to raising taxes on those people, and imposing high marginal tax rates on lower/middle income people is the poverty trap.

Nothing is going to be more efficient than a UBI because a UBI does exactly the thing it's supposed to do and nothing else. There is no inefficiency to remove.

> The only argument which favours UBI over more modest alternate welfare reforms more specifically targeted at reducing benefit withdrawal rates at the margin is the frequently-made philosophical argument that it's fundamentally unreasonable if not immoral to make handouts contingent on willingness to work and pressure them to take jobs

Really it's that if you don't have high phase out rates then you can't make it contingent on working. If two otherwise unemployed people could pay each other self-canceling payments to do each other's laundry or whatever, now they're both "employed". It's trivial to create "employment" between any number of cooperating parties that way. That doesn't happen much now because the government takes most of the "income" from that "employment" in reduction of benefits -- it invokes the poverty trap. If you eliminate the poverty trap but require employment then everyone will magically be "employed" on paper because making money creates eligibility for government benefits rather than reducing them.

And that isn't even fraud -- they really are paying each other and really are doing the thing they're being paid to do. They can even each pay each other to do the thing they each wanted to do to begin with. You don't even need another person -- create a corporation, have it pay you for whatever it is you were doing anyway and then reinvest your "wages" in the company so they have money to pay you again tomorrow. All a work requirement does is create useless inefficiency, paperwork and bureaucracy.

> I'm simply pointing out that few, if any of the people making that argument are opposed to subjecting foreigners to similar indignities if they seek to enter the country

To get in under H1B you have to be a qualified specialist. The purpose is to let in people with in-demand skills. None of those jobs are the degrading McJobs that people only take when the alternative is starvation or homelessness, and the people only qualified to do a McJob aren't "required" to do that to immigrate, they aren't eligible to immigrate at all.

> and moreover the sustainability of a UBI is entirely dependent on people not lucky enough to be citizens needing to work for a living for the foreseeable future.

It obviously isn't, because the money that funds the UBI doesn't even come from them, it comes almost entirely from other citizens in the same way that any social assistance money does.

Price and value aren't the same thing, that's part of the problem.

Suppose in response to competition the original artist would lower the price to 24c. The customer still gets the same value as before but is paying less. The customer gets 1c more of the surplus and the artist gets 1c less.

But the total surplus is still the same. The benefit the customer gets from having the art hasn't gone down, only the price.

Where this gets weird is displacement. Suppose competing art appears and now half the customers will choose the other art instead, and the customers only have time for one piece of art. Now the original art has lost half its total value because there are half as many people with it, regardless of what price was paid.

Which means it is possible to have "enough" art and more would be too much -- when you have so much that people are too busy with existing art for enough of them to choose the new art to justify its creation. But that point is well past what transaction costs will allow if payment has to be made per-customer via an ugly hack like copyright.

It may even be past the point where there is "too much" art/science/software, because you can justify over-producing a lot of junk if the result is even one more Andy Warhol or Albert Einstein or Alan Turing.

> Nothing is going to be more efficient than a UBI because a UBI does exactly the thing it's supposed to do and nothing else. There is no inefficiency to remove.

Nonsense. The former system you've described doesn't pay out to working age economically inactive people, who vastly exceed those claiming employment-linked or disability benefits in every developed country. Unlike your hypothetical example of people paying each other to do each other's laundry, these people actually exist, and needless to say transforming the system to pay out $12000 per annum to tens of millions more largely non-taxpaying people has a net cost orders of magnitude more than state bureaucracy. Most reasonable definitions of efficiency would regard it as less efficient to pay out to millions of people whose actions indicate they don't particularly want the money. (Most reasonable definitions of efficiency would also argue that it's often better to divert some of that cash towards providing additional support to people that can demonstrate a genuine need for support costing >$12000 per annum if purchased from the private sector)

So on practical grounds, it's a much, much more expensive social system than any real or feasible system with a work requirement. If you want to argue that this is justifiable because it's morally imperative that people ought to be free to choose not to work or subject to any form of state assessment unless they're foreign, feel free to make that argument. Just be aware that's the true nature of the argument you're making, and that those UBI dependents' access to the affordable consumer goods (and probably food) on a state stipend is entirely dependent on the rest of the world not being able to afford such a system to retire their own menial workers.

https://www.project-syndicate.org/commentary/basic-income-fu...