[zozbot234]

> Raising the marginal rates on the lowest earners is raising the effective rates on the lower middle class.

Those are just clawback rates. The lower middle class don't need UBI in order to pay for the necessities of life, and most UBI proposals don't expect them to be net recipients, any more than they'd be net recipients of current welfare.

> it's also lowering the effective rate on everyone above them

That's balanced by the gradual progressivity of tax rates on upper-middle incomes.

> while making sure the extra money comes from the middle rather than the top

The low rates for the lower-middle class are actually ensuring the exact opposite of that claim. The upper incomes are the source for the bulk of income redistribution in the usual optimal system as it comes out of these analyses; they just don't face prohibitive rates.

> And having the highest rates at the bottom is pretty bad incentive-wise.

It's the opposite. The bottom clawback rates apply to a smaller part of the population, that can escape them simply by earning more than the UBI breakeven point. Meanwhile the high rates there help make the whole tax schedule sustainable. It may be a counterintuitive point but it's confirmed by rigorous, automated analysis.

> marginal rate arbitrage you're reintroducing is large

The most likely response is not necessarily arbitrage, it might just be earning enough that you start paying low marginal rates after the UBI is clawed back.

[jltsiren]

> The bottom clawback rates apply to a smaller part of the population, that can escape them simply by earning more than the UBI breakeven point.

That's not true. One of the issues any serious UBI proposal needs to face are the increasing demands of the labor market. A couple of generations ago, there were still plenty of opportunities for people who had little to offer beyond a pair of hands and some work ethic. Today not so much.

The bottom end of the labor market consists of a nontrivial number of people who are not productive enough to earn a living wage. But there are still societal benefits from having them work for living, instead of being passive welfare recipients or turning to crime. To make that happen, even a low wage should increase the net income significantly above UBI.

[AnthonyMouse]

> The lower middle class don't need UBI in order to pay for the necessities of life, and most UBI proposals don't expect them to be net recipients, any more than they'd be net recipients of current welfare.

The premise of them receiving it isn't so they can buy corn, it's so they can save up a down payment on a house or have enough savings that if their car breaks down they're not completely screwed. They're supposed to get the money.

> That's balanced by the gradual progressivity of tax rates on upper-middle incomes.

That's not balancing it, it's further exacerbating it. You then have the upper middle class paying less than the full marginal rate, along with everyone above them for that part of their income range.

> The low rates for the lower-middle class are actually ensuring the exact opposite of that claim.

I feel like this is still confusing effective rates with marginal rates.

Suppose you have a 50% marginal rate up to $30,000 in income and then a 5% marginal rate up to $60,000. Alice makes $32,000/year. In this system her effective tax rate is ~47%. By contrast, Bob makes $60,000 and has an effective tax rate of 27.5%. Moreover, between the two of them they have $92k in total income of which the government gets $31600, ~34%. Why would we want Alice to have to pay ~47% instead of ~34% so that Bob, who makes almost twice as much, can pay 27.5% instead of ~34%?

Adding the UBI would lower both effective rates and result in a progressive rate curve, but why would we want to add "progressive" rates that make the system less progressive?

> The upper incomes are the source for the bulk of income redistribution in the usual optimal system as it comes out of these analyses; they just don't face prohibitive rates.

But why would you need prohibitive rates at all?

US GDP is ~$31T with a population of ~340M. To give a $12,000 UBI to every single person would be ~$4T, i.e. 12.9% of GDP, implying that's the uniform marginal tax rate you would need to collect the money. That's assuming the tax is applied uniformly (and applies to corporations as well as individuals etc.), and maybe you want to exempt non-profits from the tax or something, but as a ballpark estimate that's the number and it's not crazy high. Especially when it's a universal transfer payment and the majority of people are just getting most or all of it back immediately.

> The bottom clawback rates apply to a smaller part of the population, that can escape them simply by earning more than the UBI breakeven point.

That's not how that works. If you had a 50% marginal rate up to $30,000 and you're deciding whether to take a $32,000 job or not work, you're only getting $16,900 to put in your pocket from taking that job. Being "out of the clawback range" doesn't stop the majority of the pay from still being in it. And, of course, there are jobs that don't even pay that much, or some people might want to work part time and you're essentially screwing them for not making enough money.

> Meanwhile the high rates there help make the whole tax schedule sustainable. It may be a counterintuitive point but it's confirmed by rigorous, automated analysis.

Why do we need automated analysis to do arithmetic? You can avoid a huge increase in the rate on lower income people with a much smaller increase in a universal rate because the former only applies to a small proportion of total income and the latter applies to all of it.

> The most likely response is not necessarily arbitrage, it might just be earning enough that you start paying low marginal rates after the UBI is clawed back.

As soon as you have non-uniform marginal rates, arbitrage is going to happen. We don't have to speculate here because we currently have non-uniform marginal rates and it's extremely common. Family members in a lower tax bracket get added to the payroll to soak up taxable income at the lower rate.

[zozbot234]

> They're supposed to get the money.

We're talking about the lower middle class, they can sustain themselves by working, essentially by definition. They can get the money (compared to the status quo) - by working, because they'll be paying negligible overall rates on their own earnings as soon as the UBI is fully clawed back. Redistribution (UBI) is costly (being funded by high rates on the higher incomes, that deter them from earning more), so in most practical proposals it is largely limited to those who really are unable (or perhaps unwilling, but in a near-optimal system such 'unwillingness' is most often explained by practical inability) to earn enough for a tolerable life. Giving any part of the middle class a net-UBI only to claw it back with high rates (as it must be at some point in the tax schedule) wouldn't leave them any better off in practice and it would set up pathological anti-work incentives. It's a pretty clear non-starter.

> You then have the upper middle class paying less than the full marginal rate

If you're talking about effective rates being less than marginal rates, that's just inherent to any progressive taxation. There is in fact a strange theoretical result that the very top earner should ultimately be facing a zero marginal rate on their very last dollar of income since any non-zero rate on that "top" level is pure overhead, but of course that's a bit of a curiosity and hard to apply in practice, except maybe as a broad caution against the common idea of levying punitive rates on the very highest incomes. Progressive taxation at the high end is still the practical optimum.

> I feel like this is still confusing effective rates with marginal rates.

I feel like you're disregarding the fact that the high-rates at the bottom are purely UBI-clawback rates; in fact, $30,000 per year seems quite infeasible as a UBI range, so ISTM that you're inadvertently getting the wrong idea entirely about what the system might imply!

[AnthonyMouse]

> We're talking about the lower middle class, they can sustain themselves by working, essentially by definition.

Then why does the existing system provide all manner of healthcare subsidies, student loan subsidies, etc. with phase outs extending well into that income range? The ACA subsidy for a family of four isn't fully phased out until $132k.

> Redistribution (UBI) is costly (being funded by high rates on the higher incomes, that deter them from earning more)

Aggressive claw backs are more costly because they require much higher marginal rates on lower income people and it's high marginal rates that provide the disincentive to work. Even before considering the argument that people being paid low hourly rates are easier to deter to begin with, because paying someone else to do something for you in order to free up time to work for money instead of doing it yourself is more costly to someone who gets a lower hourly wage.

> Giving any part of the middle class a net-UBI only to claw it back with high rates (as it must be at some point in the tax schedule) wouldn't leave them any better off in practice and it would set up pathological anti-work incentives.

Have you done the math on this one? The bottom 50% of people have <15% of total personal income. You get less money by increasing their pre-UBI effective rate by 50% than by increasing the universal effective rate by 7.5%, while providing them with a massive disincentive to work.

> If you're talking about effective rates being less than marginal rates, that's just inherent to any progressive taxation.

No, we need to distinguish two things here. One is the incremental effective rate for taking a job, i.e. if the employer pays $X to employ you, what total percent of X goes to the government? This is the thing that matters when choosing whether to take the job (incentive to work). You want this number to be uniform because any decrease in one place requires an increase for someone else and you lose jobs whenever you cause the rate to exceed the surplus derived from the job, so combinations of high and low rates are less efficient than uniformly medium ones. You especially don't want it to be higher for lower income people, because then the increase on them is disproportionately large.

The other is the effective rate net of unconditional benefits. When you get benefit (i.e. UBI) regardless of whether you take the job then that's included here but not there.

You get the same effect as a progressive rate structure in the second sense with a flat rate plus a UBI, because everyone pays the same marginal rate and therefore the same effective rate in the first sense, but someone who makes $54,000, pays a third of it in tax but receives a $12,000 UBI is on net paying $4000, i.e. 7.5% rather than 33% of $54k, whereas someone who makes a billion dollars is still paying a third and receiving the UBI, but $12,000 is a trivial percentage of a billion so their effective rate is still ~33%.

> There is in fact a strange theoretical result that the very top earner should ultimately be facing a zero marginal rate on their very last dollar of income since any non-zero rate on that "top" level is pure overhead

This theory assumes that the primary decision here is whether to work an additional hour, but the primary decision is actually which job to take, if any. You might make three times as much if you become a doctor, but then you give up 8 years of earnings to go to school and exit that with six figures in student debt. If you're deciding when to retire, or whether to enter the workforce but then have to pay much of your earnings in childcare and commuting expenses, you're either keeping/taking the job or you're not. The "incremental" rate when making the decision applies to the majority of or entirety of the compensation, not just the last dollar.

And notice that you can simultaneously screw this up on both ends. If you put oppressively high rates at the low end, even if that only lowers the high end rates by a minor amount, it just became relatively a lot more attractive to make three times as much money. Then you get elite overproduction and cost disease because you're screwing the janitors and service workers so hard on the tax rate. Some of them fall out of the workforce, others turn to diploma mills and employers have trouble hiring in that income range for necessary jobs and then have to raise prices on everyone. It's the same terrible thing the existing system does and what we should be trying to prevent.

> I feel like you're disregarding the fact that the high-rates at the bottom are purely UBI-clawback rates

There aren't supposed to be clawback rates. That's the entire point -- to get rid of the punitively high work-disincentivizing marginal rates for lower income people present in the existing system.

> in fact, $30,000 per year seems quite infeasible as a UBI range

That number is lower than the existing income level where people pay zero/negative federal income taxes net of benefits.