Some commenters on my recent post on the Universal Basic Income (UBI) seemed to claim that there would be no big effect on most people because the UBI they received would be taxed back. So, according to this argument, the net effect on most people would be approximately zero. But some basic economics says that that’s false.
Imagine the government gives you $10,000 per year but tells you that once your overall income reaches $40,000, you will lose $1 of that $10,000 for every additional two dollars you earn. That means that on top of your regular marginal tax rate you pay (which includes the federal income tax rate, the state income tax rate, and the payroll tax rate), you will pay an implicit tax rate of 50%. That will have a strong disincentive effect.
Moreover, one of the ways economists think about incremental taxes is to divide the effect into an income effect and a substitution effect. The income effect is that additional taxes make you poorer and so you “demand” less leisure: you work harder. The substitution effect is that the price of leisure has fallen and so you “demand” more leisure: you work less. Which outweighs which is an empirical issue. For married women, the substitution effect typically is much stronger than the income effect, so they work less when marginal tax rates rise.
But the UBI introduces a new wrinkle. The $10,000 that everyone gets increases people’s real income and, therefore, increases their demand for leisure.
So the overall effect of the UBI that is phased out is made up of three components: (1) the income effect due to the UBI which reduces work: (2) the income effect due to the phasing out, which increases work; and (3) the substitution effect due to the phasing out, which strongly reduces work when the phaseout is on the order of $1 lost for every $2 earned.
A way to reduce this strong disincentive effect is to phase out the UBI more slowly. For example, the government could reduce the UBI by $1 for every $3 earned past a total income of $40,000. Then the added implicit marginal tax rate would be “only” 33.3 percent. This has two added problems, though. First, it makes the UBI more expensive than otherwise because everyone whose total income is between $50,000 and $80,000 nets some portion of a UBI. Second, it subjects a much wider swath of people to the implicit marginal tax rate.
Giving people money and then taxing it back as you earn more is particularly perverse tax policy.
READER COMMENTS
Todd Ramsey
Nov 12 2023 at 10:37am
What if the UBI replaced the current patchwork of means-tested benefits like food stamps, housing subsidies, heath care, AFDC, etc?
Currently, as is sometimes mentioned in EconLog, the combination of those support programs result effectively in a marginal tax rate of 100% for recipients contemplating work.
Wouldn’t replacing existing benefit programs with UBI distort the work-leisure tradeoff much less than our current policies?
Dennis Senger
Nov 12 2023 at 11:04am
While I agree that the UBI would be less distorting for those who would otherwise receive means-tested benefits, it would apply to everyone, so a majority would have worse incentives.
Matthias
Nov 12 2023 at 6:35pm
Why would the majority of worse incentives?
Assume for a moment that you abolish (nearly) all other benefits and design your UBI in such a way that both the net payments and the net marginal payments for everyone stay (nearly) the same as before.
By net payments I mean tax minus benefits. Be that UBI or in-kind benefits like toll free road usage or subsidised flood insurance for your beach front property or dole money, or not including imputed rent on owner occupied property in the income tax.
The incentives are largely the same as before. (Only that instead of only care owners benefitting from free roads, everyone benefits from UBI. So car owners will presumably get a bit less money than used to save in tolls, and non car owners will see a net increase in their income. Same for home owners vs renters.)
vince
Nov 12 2023 at 1:06pm
An obvious alternative to phasing out UBI at higher incomes is simply including it in taxable income. Then its progressivity is the same as other taxable income.
Complete phase-outs, over an income window, are sometimes called a tax bubble. An example is Social Security. Its taxability quickly goes from 0 to 85 percent.
Matthias
Nov 12 2023 at 6:45pm
Yes, phase outs are a bit silly, since you can achieve the same result with income tax brackets. And the UBI doesn’t even need to be included in the taxable income: no matter whether you include it or not, you can adjust the tax brackets to achieve the same effect.
However what David says is still true: if you hold the total economy wide tax take of your (effective) income tax constant then less tax or even negative tax at a lower bracket means a higher marginal rate in at least some of the higher brackets.
Whether your use phaseouts or not or whether include the UBI in the taxable income or not, doesn’t make any conceptual difference: you can adjust the brackets.
You can escape this conclusion, if you find money from elsewhere. Eg a land value tax (or even a property tax) could be an alternative source of tax take that’s not tied to income.
robc
Nov 12 2023 at 3:28pm
The only way a UBI makes any sense to me is if it is combination with a flat tax. Then theere is no “tax bubble” or “phase out” or whatever, the marginal tax rate is the same all along the spectrum, for net receipients to net payers.
The problem (a problem, probably not THE problem) is that it would have to be at least 35% to make any sense, which combined with state/local/payroll/etc taxes puts the marginal tax rate for everyone very high.
Matthias
Nov 12 2023 at 6:19pm
You can combine a phase out and income tax brackets to get any desired effective income tax you like.
I don’t really know why anyone would want a phase out, though. That’s just unnecessary complication, and belies the U in UBI. The phase out doesn’t give you anything in terms of policy you can’t get from tax bracket design.
andy weintraub
Nov 12 2023 at 5:08pm
But what if it’s taxed back at the marginal rate appropriate for the total taxable income, which would include the UBI?
I believe that’s what Milton Friedman had in mind when he suggested a “negative income tax”.
Matthias
Nov 12 2023 at 6:25pm
It doesn’t really matter, if you tax the UBI: taxing the UBI is equivalent to paying slightly less (untaxed) UBI and changing the income tax brackets slightly.
What David says is right: if you give people more money in the lower tax brackets, you need a higher marginal rate at a higher tax bracket. (Unless you have other sources or revenue, of course. Like a land value tax. Or because you increase efficiency by removing the bewildering array of complicated benefits etc.)
Daniel
Nov 13 2023 at 7:36am
@Matthias, I suspect you, like I, groan every time a politician says they’re going to simplify the tax code by having fewer brackets. If there’s one thing we’ve learned from implicit marginal rates, it’s that we want more/smoother phaseouts/brackets, not fewer/steeper!
Matthias
Nov 13 2023 at 9:07am
Well, I moved to my adopted home of Singapore a few years ago. Our tax system is simple and arguably provides good value for money (in terms of what I got in terms of government services for my tax dollars).
At least the bang-for-my-buck ratio is much better than it would have been in my native Germany.
Though funny enough, I am paying more tax here in Singapore that I would probably ever have done in Germany. In absolute terms that is: salaries are so much higher here that they more than make up for the lower tax rates. But I’m obviously ok with the trade-off.
Thomas L Hutcheson
Nov 13 2023 at 8:24am
Bottom line. EITC/wage subsidy is better than UBI.
And either would be less worrisome if financed with progressive consumption taxes rather than progressive income taxes shot through with deductions instead of tax credits for items of favored consumption.
Jose Pablo
Nov 13 2023 at 1:07pm
Don’t phase out the UBI at all. You only need a “two line” tax code:
And IVA of around 30-35% on every expenditure incurred (which would mean around $25,000 per year in taxes for the average household, approx $75,000 on expenditures)
A UBI of $4,000 a year per person (which means around $10,000 per year for the average household of 2.6 people)
Households spending $33,000 per year would be tax free. Household spending less will have a negative tax rate.
You can still finance (although you don’t need it) a “20% of GDP spending” government* (even bigger, not advised, if you included property taxes, Pigouvian taxes and government owned infrastructure tolls…)
No income taxes, no payroll taxes, nothing … and a pretty progressive system: households spending $250,000 a year will have a 30% tax rate compare with the 13% effective tax rate of the median household (spending $50,000 per year) and a negative effective tax rate for around 40% or more of the population.
And you can get big savings by firing most of the IRS agents, passing out all welfare programs and firing all the people managing the welfare state.
And also get a great increase in productivity by making most of accountants and tax advisor doing something actually beneficial to society with their bright minds; instead of wasting their lives going thru extremely boring and ever increasingly complex tax codes.
Bob Bell
Nov 13 2023 at 9:04pm
Brilliant, straightforward and effective. So of course it ain’t got a snowball’s chance of ever happening.
RPLong
Nov 16 2023 at 9:17am
This is a large overhaul to the existing tax/welfare system, too drastic to implement all at once, there are too many political barriers to such a thing. So this is an interesting hypothetical proposition.
Which, to me, raises the question of why – if we’re considering hypothetical, ostensibly libertarian proposals – we wouldn’t just drastically reduce government expenditure and taxes across the board. What is the benefit of including a UBI in this kind of hypothetical? Are you really just pleased with the idea of a UBI in theory?
johnson85
Nov 17 2023 at 4:57pm
I think for most libertarians (or libertarianish), the UBI is appealing because it is a more efficient and less destructive form of the welfare state. It’s more politically feasible than just getting rid of the welfare state, so it’s an alternative to getting rid of the welfare state.
You are correct about it being basically equally politically impossible, but if you believe in ranking impossible policy proposals as far as just how impossible a proposal is, it’s less politically impossible than just getting rid of the welfare state.
If you do a universal basic income that doesn’t do away with Social security and medicaid (or at least phases it out for people not old enough to vote now), then it becomes much less impossible (while still more or less being impossible) but of course the math doesn’t come anywhere close to working at that point.
Comments are closed.