Comment by everythingswan | Hacker News Reader

Comment by everythingswan | Hacker News Reader

Parent Full thread

I see a straw man here. It seems like you can argue against the theory or argue against the actual implementation. The latter was chosen...but with made-up parameters that make it easy to defeat. Examples

"...a UBI of just $1,000 per month would cost around $4 trillion per year, which is close to the entire federal budget in 2018. Without major cost savings, US federal tax revenue would have to be doubled, which would impose massive distortionary costs on the economy..."

So we need major cost savings or taxes to go up.

"Sacrificing all other social programs for the sake of a UBI is a terrible idea. Such programs exist to address specific problems, such as the vulnerability of the elderly, children, and disabled people. Imagine living in a society where children still go hungry, and where those with severe health conditions are deprived of adequate care, because all the tax revenue has gone to sending monthly checks to every citizen, millionaires and billionaires included."

This came our of nowhere as social programs were not mentioned before this line. It's basically fearmongering at this point.

Maybe someone can help me understand what the argument here is. Perhaps I'm being harsh on the author but I just don't see how you can make this argument with genuine intentions.

> This came our of nowhere as social programs were not mentioned before this line. It's basically fearmongering at this point.

Perhaps he did not spell it out explicitly, but this is what Andrew Yang's policy proposal is. Since he is the leading candidate on UBI at this point, it makes sense to address how he plans on handling it.

*was. He amended his proposal such that UBI is additive to current social programs.

What makes you say that? His website lists current social programs as the first source of money for this program. People either switch over and see a marginal benefit, or they do not switch and they get nothing from the Freedom Dividend:

> The means to pay for the Freedom Dividend will come from 4 sources:

> 1. Current spending. We currently spend between $500 and $600 billion a year on welfare programs, food stamps, disability and the like. This reduces the cost of the Freedom Dividend because people already receiving benefits would have a choice but would be ineligible to receive the full $1,000 in addition to current benefits.

https://www.yang2020.com/what-is-freedom-dividend-faq/

It gives the Freedom Dividend this weird behavior of providing the least help to people who need it most.

This felt like a purely semantic argument to me; as he is actually suggesting we do something rather similar to UBI:

"Besides, a more sensible policy is already on offer: a negative income tax, or what is sometimes called “guaranteed basic income.” Rather than giving everyone $1,000 per month, a guaranteed-income program would offer transfers only to individuals whose monthly income is below $1,000, thereby coming in at a mere fraction of a UBI’s cost."

It's a substantially different thing with quite different incentives.

For a particular example, someone receiving guaranteed basic income is in some sense "stuck" as a noncontributing member of society as it makes no sense for them to start working e.g. a part-time job since whatever money they earn would simply reduce their GBI payment with no benefit to them.

GBI helps solve some problems, but it's not really a solution to the type of problems that UBI wants to solve, it's something quite different.

The GBI would have to decline gradually as your income rose, not all at once, as marginal income tax brackets do now.

Say it gives 1000 a month to someone earning 0 a month, and grants 500 a month to someone earning 1000 a month, and nothing to someone earning 2000 a month (shortly followed by the positive tax rates for income beyond that).

>The GBI would have to decline gradually as your income rose, not all at once, as marginal income tax brackets do now.

By itself it works, but not when combined with other programs.

For example, take someone on 2 different forms of help. Both give $1000 a month, but for every $1 you earn, they decrease benefits by $.50. So each slopes off at 50%.

So a person gets a part time job making $800 a month.

The first program cuts its benefits by half of that value, giving only $600. The second program does the same. Either way they take home $2000, meaning they earn nothing for working part time.

Now, this is a simple scenario and can be fixed by making the decline very gradual in this case. But it shows one immediate problem, this solution doesn't apply at scale. So as long as you have enough sources of benefits the decline still ends up making working not worth it. Adding in taxes and private charity can end up complicating this further but the general result is that it produces local maximums that act like a trap.

Also, if you fix this my making each one decline slower, you end up with it not being as easy to sell because of how high it kicks in. For example, instead of 50% decline, making it a 20% decline. So someone making $4,500 a month still gets $100 in benefits (1000 - .2 * 4500). That can be a hard sell politically regardless of how much benefit it is because it allows people to make the true claim that those earning $4900 a month still benefit from this benefit, despite it being only $20 a month.

There is another complication with marriage and how it impacts this benefit (and if that creates an incentive for or against marrying).

Lastly, even if you do design a system that works well enough, will the public even understand it? Look at how many people, even people who have jobs dealing with money, struggle with tax brackets.

As such, I think it seems much easier to just say everyone gets $1000 (though there are still other questions, such as anything involving children).

That would work, but it makes the GBI much more expensive. If you did the decline gradually enough such that there was no effective disincentive for 99% of the population, you'd have something that looks and costs very similar to a UBI.

You misunderstand tax brackets. You don't get taxed on your entire income for a higher rate if you get bumped into a new bracket, only the income that is over the threshold.

This creates an issue where jobs that pay only slightly more than $1k a month end up not being worth it. For example, someone working 30 hours a week at $10 an hour would see their income only go up ~$200. You could phase it, but that only weakens and does not remove the underlying issue and there are already existing cases where multiple problems with phasing stack, ending up in the same problem. Going with a simpler form of UBI that gives everyone $1k no matter what completely eliminates the issue (well, there is a minor issue given that money have a diminishing returns in value, but the impact of this is so far weaker than the issue of cut offs or phasing out benefits).

I disagree that someone would see it as "only" $200 a month. That's a 20% pay increase for 30 hours a week - hardly onerous.

I mean, incentives 101 tell me lots of people would be happy to work 30 hours a week for extra cash.

I mean, aren't they already not worth it? A job paying $10 an hour barely puts you above the poverty line even working full time.

You can just double all the numbers used if you like, as $2000 UBI isn't some outlandish suggestion.

It's economically quite different. Much less is spent, but lower income people end up facing very steep marginal tax rates. If you make 12k or under, you might as well not work or work in an unpaid role. Even if you make more than that, given a choice of working 40 hours/week for 24k or 0 hours/week for 12k, many or most people will go for the 12k. With a UBI, you're paid the 12k, and if the job's salary is 24k, you get paid 36k total, which is a much more attractive proposition. A similar effect keeps some SSDI recipients in a poverty trap.

All that said, there are lots of nonlinear effects, and any solution will have to be phased in slowly to account for and respond to them. There are also ways to make the guaranteed income have much better marginal effects, though doing so also makes it more expensive. In fact, those mitigations if taken to an extreme lead to a UBI.

Yeah, that basically is semantic. UBI would achieve the same thing, but, instead of negative tax, would just credit everyone and then alter the tax rates to achieve equilibrium.

Personally, the improved efficiency from simply collapsing the existing apparatus which is focused on preventing people from cheating seems likely to work. If the goal of the program was to fairly allocate a resource, rather than to prevent people from cheating it would consume a lot less overhead. that is where you can actually find some of the efficiency that the author wants.

Honestly, if you cut a check to a millionaire for $1000, and then tax that person on that $1000, it doesn't work that differently from the fact that we get taxed on the value of our tax returns. it ends up being recursive anyway.

Isn't UBI and negative income tax mathematically equivalent?

Let's say you have $1000 UBI and 20% income tax. Then the money you end up with after UBI and taxes is 0.8x+1000 dollars.

That's the same as paying 20% (signed) income tax above $5000 where you end up with x - (x-5000)*0.2 = 0.8x+1000 dollars.

> This came our of nowhere as social programs were not mentioned before this line.

Wasn't the entire prior paragraph his justification for it?

> Perhaps I'm being harsh on the author but I just don't see how you can make this argument with genuine intentions.

I come at this entirely from the other side of the political spectrum, I think his facts are wrong and his ideology worse, but he appears entirely sincere to me.

> It's basically fearmongering at this point.

All arguments that a policy will entail bad results are logically equivalent to fearmongering. You're just saying you doubt his intentions using another synonym.

Reply on news.ycombinator.com