[This post was requested on Patreon by Daniel, who asked me to write up an argument against guaranteed basic income. Each month, one backer at the $5/month or above level is randomly selected to suggest a blog post or story topic. You may find my patreon here.]
The reason one should not support a guaranteed basic income is that economists don’t support it. In fact, weighted by the economists’ confidence in their opinions, 84% of economists surveyed either disagree or strongly disagree with granting every American citizen over the age of 21 a guaranteed basic income of $13,000/year, financed by eliminating all transfer programs.
…wait, you mean I have to write more than that? Come on.
As always, I’m not an economist, but looking at common themes in the explanations the economists of the IGM Economic Experts Panel, I think I can come up with some more answers.
The GBI is expensive. Really, really expensive. The GBI proposed in the question economists were polled about would cost about three trillion dollars a year, equivalent to all the tax revenue of the United States federal government. It is generally believed, even by the most libertarian among us, that countries should continue to have things like “an army” and “diplomats” and “a place for the President to live.” Most of the liberals I know would go so far as to say that we should have environmental conservation, publicly funded science, and foreign aid. A GBI is either going to lead to a massive increase in taxes or a cut in a lot of really basic things.
And the benefits aren’t actually that great. Thirteen thousand dollars for each adult is enough to get most families over the poverty line (although not single parents with two or more children). Living at the poverty line is pretty terrible, but it’s not the worst possible thing; you don’t starve. Of course, most people who are at or below the poverty line benefit from food stamps, Medicaid, and other programs intended to help the poor, which don’t exist anymore because we slashed them to fund the guaranteed basic income. So we’d expect that being a poor person would become significantly harder.
I’m talking about one specific proposal, but these arguments apply to pretty much every level of guaranteed basic income. You can give everyone $20,000, which is enough to live on happily in many parts of the country, but then you’d better find some more sources of tax revenue. You can give everyone $5,000 and be able to afford the continued existence of the White House, but then it’s not going to do a hell of a lot of good for most poor people.
Part of the reason the GBI is so expensive is that it goes to everyone. Bill Gates gets a $13,000 check; I get a $13,000 check; the homeless guy down at the bridge gets a $13,000 check. This is often justified by pointing out that means-tested programs lead to disincentives to work. If working to earn a thousand more dollars will lose you your food stamps, and you have to pay taxes on the extra thousand dollars, you can quickly wind up with more than half of your additional income going to the government (either in the form of taxes or in the form of means-tested programs you’re not benefiting from). In some cases, nearly 100% of your additional income would go to the government.
That’s a serious problem with means-tested programs. However, there has got to be a better way to solve it than giving everyone in the United States $13,000. For instance, we could phase out benefits more slowly. We could also work on simplifying the absurdly complicated array of social welfare programs. A lot of times, no one intends to make someone have a 80% marginal tax rate; it’s just that when you have dozens of programs affecting poor people, it’s impossible to account for all the ways they could interact with each other.
Related to targeting the most vulnerable: a guaranteed basic income does not account for the fact that severely disabled people are more expensive than other people. For instance, consider a quadriplegic. She is currently unable to work, so she lives on her $13,000 guaranteed basic income check. Let’s say she requires a home health care aide to bathe, dress, and use the bathroom, which costs her $19 an hour for two hours a day of care. Over the course of a year, this adds up to $13,870– which is more than her basic income check, leaving her with zero money for food, rent, entertainment, or health care.
Right now, home health care aides– as well as innumerable other services required by the severely disabled– are covered by Medicaid. If we eliminate Medicaid, then many– perhaps most– severely disabled people will be unable to afford the care that allows them to live. Some will be taken care of by friends or family, or will be rich enough to afford it themselves, or will go into the kind of nursing home you can afford for $13,000 a year (spoiler: it is a really terrible nursing home). Many will die.
Many of my readers, I know, basically think that severely disabled people should not exist. That is another conversation for another time. But let me point this out to those readers: most quadriplegic people aren’t born quadriplegic. You could walk out of your house today and get hit by a car and wind up quadriplegic. So the question is whether, if that happened, you would like to continue to be able to live independently, or you would like to die.
Finally, the idea of a guaranteed basic income does not necessarily work well with increased immigration. If the guaranteed basic income is available to everyone, there is an incentive to come to the United States and not work. If the guaranteed basic income is not available to everyone, we create a permanent underclass. Neither is satisfactory.
Thing of Things 
Calling immigrants excluded from the welfare state an “underclass” is the non-central fallacy, at best. Excluding them from the country is treating them as much more of an underclass than not giving them free stuff.
LikeLike
mdaniels has been banned for consistent lack of interesting or insightful opinions.
LikeLike
“In some cases, nearly 100% of your additional income would go to the government.”
It can be worse than that with welfare cliffs. The exact numbers vary with circumstances, but going from $30k gross income to $40k can result in a loss of more than $20k in net income + benefits. It’s a 300% marginal tax rate.
LikeLiked by 2 people
Damn, I was waiting for the ITT answers…
LikeLiked by 1 person
In case anyone’s interested, here’s a link to a classic defense of basic income, along with a ton of responses from other scholars (some supportive, some critical) and brief reply by the author, Philippe Van Parijs. Some of the objections are very similar to Ozy’s.
http://bostonreview.net/forum/ubi-van-parijs
And I’m gonna shamelessly plug my university’s upcoming conference on basic income. Van Parijs, author of the above article, will be a keynote speaker.
https://www.bgsu.edu/arts-and-sciences/philosophy/workshops-and-conferences/the-future-of-work–technology-and-a-basic-income.html
LikeLiked by 1 person
Thanks!
FWIW, I found Van Parijs’ response to “won’t it cost too much” to be heavy on words and short on math. A paragraph like Ozy’s, that provided example numbers for how much we might pay each recipient, and how we would fund it would be much more helpful to me.
Ozy has made a good case that $13K/person/year is (a) not very much money on a personal and (b) very expensive nationally – I’d love to see someone work the numbers for a pro-UBI side, instead of 3 paragraphs of “well, some people say it would cost too much, and it would certainly cost a lot, but the situation is very complex.”
FWIW, my concerns are:
1) It seems like it would cost more than we are willing to pay, or would be so small that it wouldn’t justify withdrawing existing social programs.
2) If we pay for children, and it is possible to raise a child on less than we pay, I am somewhat worried about the marginal incentives on early and low-investment child birth.
3) I’m also worried about the incentive to work, and on immigration.
4) Setting the number is hard too. Do we want people to work at crappy entry level jobs and work their way up, or have we agreed that no one should work those jobs? If we set the number low enough that people still have an incentive to work, then we’ll feel bad about the people who honestly can’t work. If we set it higher, then the best case is that the effect on entry level jobs will be inflationary, which means inflation will eat away at the benefit or that we’ll be on a spiral as the benefit keeps increasing to chase inflation, and there will be fewer entry level jobs for anyone to do.
LikeLiked by 2 people
People are going to have an incentive to work pretty much no matter what, because pay differentials won’t go away. It might become cheaper to automate eg fast food ordering, but is that necessarily a problem?
LikeLike
@gazeboist
People don’t automatically get to benefit from pay differentials, though. The average road side worker doesn’t seem able to become a programmer, for example.
But we still need road side workers. If the UBI reduces the number of people willing to do that work, pay has to go up for jobs like that, while pay for enjoyable jobs will drop even more. This is a market distorting effect, where I’m wary about the consequences (I have never seen a serious simulation/analysis of the effects that one may expect).
LikeLike
1. Is ‘remove Medicaid’ a common UBI position, at least within the greater complex of ‘remove now-superfluous programs’? I tend to just work under the assumption the UBI/Job Guarantee Utopia is also the Socialized Healthcare Utopia. (I live in a country that compared to the US is in fact the Socialized Healthcare Utopia, so I tend to kind of forget it’s a highly controversial topic there.)
2. I am sure there is a simple and easy solution to this problem. However, every simple and easy solution I can think of is…not in the Overton window. (dysphemistically: extremely racist)
LikeLike
(slight addendum: I am not racist, do not support UBI policies that would discriminate by race, etc.)
LikeLike
actual addendum: goddamnit, I think too fast. ‘2.’ is a response to the bit about immigration.
LikeLike
It depends on whether people think it would become superfluous? My personal opinion is that parts of it would become superfluous (ie the parts that pay for some sort of standard maintenance healthcare like flu shots and regular checkups, and maybe instant catastrophic insurance), but other parts would not (medical care for people who have unusually high long-term care costs, basically).
That second group would still need some kind of medicare/medicaid equivalent, but I’m not sure how that would compare with current costs.
LikeLike
> Part of the reason the GBI is so expensive is that it goes to everyone. Bill Gates gets a $13,000 check; I get a $13,000 check; the homeless guy down at the bridge gets a $13,000 check.
At worst, bringing up Bill Gates is an appeal to emotion that has no business influencing policy. Giving a few billionaires 13k each is a meaningless drop in the ocean. Even if you restrict yourself to, say, “the 1%”, we’re very literally talking about 1% of the costs of the program. The vast majority of the expense is from everyone else.
At best, bringing up Bill Gates is a response to some wimpy pro-UBI positions, and I’d rather hear your response to a more sensible plan. Consider a flat income tax rate (e.g. 45%) alongside a UBI of $13k/year. In this case, for the very poor, this system is a simplification of the current welfare system that isn’t so gameable and doesn’t have stupid marginal rates at any point. (The marginal tax rate under this plan is always 45%.) For everyone else, we’re NOT needlessly giving people $13k. Instead, that $13k grant is an implementation detail of a progressive tax system, serving the same purpose as the standard deduction and the complicated schedule of marginal tax rates that we have now. If the average person ends up netting the same amount in UBI – taxes, saying that we’re wasting money to give them the UBI would be disingenuous.
LikeLike
Thanks – Would a flat 45% income tax + removal of most transfer payments pay for a $13K UBI?
LikeLike
This ends up at slightly more complicated but probably easier to implement programs like a progressive income tax that just goes negative at one end.
LikeLike
UBI is a philosophical idea. Negative income tax is a possible implementation of that idea. I don’t think comparing them to each other is meaningful; they are the same thing viewed at different levels of specificity.
LikeLike
This. Any sane UBI proposal I’ve heard also eliminates progressive taxation (as that would be implicit with UBI anyway). Either with a flat income tax, or by replacing income tax with a much more significant sales tax.
In both cases, the effective tax rate for most people would stay roughly the same, but with a more elegant system, without any “steps” in tax rate that provide a disincentive to earn more, and with much less opportunities for gaming the system and much less need for bureaucracy to counteract said gaming.
You’d still need to provide for disabled/ill people with exceptionally high needs, but that could be done via an insurance system that’s paid for by part of the UBI.
Also, it should only be available to citizens, and immigrants who have paid taxes for a substantial time (maybe 5-10 years; they could even get the UBI as a tax reduction before that, just never a negative tax rate).
LikeLike
UBI would be financed by increasing State expenditures relative to GDP which is very low in the US. You don’t have to slash other programs.
Anyway, The “UBI is too expensive” argument stems from a confusion between costs and transfers. UBI moves huge amounts of money but most of this flow is then taxed again so that the money can be spent on other programs. These programs are a real cost in the sense that they provide real services like healthcare or education, they mobilize ressources that could have been used for something else.
UBI isn’t spent on anything in particular, it just rearranges the distribution of income in the economy by decreasing the importance of market income in favor of after tax income. More money is flowing through the government but the government isn’t actually using it, it merely redistributes it. It’s a huge transfer but it’s not a cost.
LikeLike
Does this change how hopeful we should be about GiveDirectly’s UBI experiment?
LikeLike
> equivalent to all the tax revenue of the United States federal government
I have to argument about comparing it to the _federal_ revenue. Many of the savings would be on the state level, so the relevant comparison is with the total revenue of all states plus the federal government.
LikeLike
Have there been other polls of economists on UBI? I’d like to see broader questions like “in your conception of ideal US welfare state, a UBI would {an important,the primary} form of social support”, or “a significant amount of social spending would be better replaced by a UBI”. A battery of variations like that would give us a clearer picture of where the profession stands on the general concept.
LikeLike
“Many of my readers, I know, basically think that severely disabled people should not exist”
Wait what?
LikeLiked by 2 people
My thoughts about UBI are complicated.
1. People who need MORE support in a variety of forms should have that. I think most people imagining otherwise literally haven’t thought about it. But even if those supports exist, there’s still much much less bureaucracy for many other existing benefits. I know some people genuinely want a policy which harms those people, but I hope that’s few 😦
2. The only sensible interpretation from my PoV is corresponding tax increase that leaves everyone earning more than a little above the basic income cut-off as well off as before or worse off. From my PoV, that’s the POINT, a simplification, and a bonus to the people who need it, the people scrabbling from job to job, or bankrupted by a bit of bad luck, not to make middle class better off. I think a simultaneous redistribution (by increasing tax on higher income a lot, or by inflation) might well be a good thing, but I don’t think they’re the same, even people often conflate them.
I think even taking those into account, UBI might *still* be unworkable, but I’d prefer to see objections to what seems to me the most plausible version, not the loudest-noised version.
I don’t have a clear argument here, but I think things like, with fire service, the government provides, and it works best if the government just provides it to everyone, rather than using a complicated rationing service to judge how much people need. The same for transport. The same for police (which often falls short). The same for healthcare (in many countries). Ditto education (ideally education that’s actually useful). To me, that seems a trend which could easily be extended to other needs people are likely to have, like being tided over from one job to another, or having the liberty to study towards a more useful job, rather than having to take one immediately to survive.
Another way to look at it would be, unemployment and disability benefits should be generous and err on the side of assuming that *most* people are honest, and benefit from them, and making people less terrified of ruining their whole life because they needed support and it was withheld, would on balance be a big benefit, not drain, to the economy.
But I know it’s far from that simple. I am also really uncertain about doing it for one country if you don’t do it for the whole world (but then, you could say the same about ANY money spent on the country you’re in, if you’re spending it anyway, it might as well be efficient).
LikeLike
The critiques I have seen of UBI that appeal to how much it costs, and to the fact that it gives money to the wealthy as well as the poor, including Ozy’s, seem to me to be making an elementary error.
Take the following policy: We send a check to everyone for $1000, financed by raising everyone’s taxes by $1000. This would be a silly policy. But it is not a silly policy because of how expensive it is, or because of the magnitude of tax increase required to finance it. It is exactly as silly as a policy which sends a check to everyone for $2000, financed by raising everyone’s taxes by $2000. It’s equivalent to doing nothing at all. Nominally, the first policy “costs” 300 billion dollars, and the second one “costs” 600 billion dollars (assuming 300 million people). But in any sense that matters, the second policy is not twice as costly as the first. The “costlier” program is exactly equivalent in impact as a less “costly” one (which is to say, no impact at all).
But the “three trillion dollars!” cost of the UBI is calculated in the same flatfooted way.
And just as in the examples above, the UBI is equivalent in impact to a less “costly” policy.
Compare the following two policies: first, a simple UBI that gives $13000 to each person, financed by a progressive tax.
Second, a policy that works like this: first, take 13000, and subtract the amount that the individual would have paid in tax under the first policy. If it is positive, send them a check for that amount. If it is negative, then tax them that amount.
This second policy is massively less “costly” in the way Ozy has calculated the “cost” of UBI. In addition, it does not send a check to rich people as well as poor people – it only sends a check to lower income people, and taxes higher income people. Wow! If those were genuine reasons why UBI was bad, this policy seems way better!
But of course, the second policy is just a very poorly disguised notational variant of the simple UBI. Any genuine reason why UBI is bad must apply to this policy as well.
LikeLike
Can you work the math on how much a $13,000 per person UBI would cost?
LikeLike
That depends on a few things, including how we’re going to understand “cost”. The point above was that the mere nominal cost of the policy was a really bad way of understanding “cost”, since it turns out to be completely irrelevant to how good a policy is (functionally identical policies turn out to have different costs).
On one way of looking at it, since UBI is just a wealth transfer, and each dollar taken away from one person is given to another, the cost is just whatever negative incentive effects there are from the transfer (the loss from people working less hard).
Another, perhaps more natural, way of looking at it is that the cost is what you get when you take all of the people who pay more than they get from the policy, and add up the difference. This is better than the way Ozy was calculating it, since it assigns the same cost to the equivalent programs I mentioned. How much the cost is, in this sense, will depend on the tax used to finance it. If the 13000 handout to everyone is financed by a 13000 tax on everyone, the cost would be zero. As you get more progressive, the benefits (which accrue to the poor) go up, and the cost (which accrues to the rich) goes up. Just to illustrate how it goes, suppose we finance it with a silly tax which goes like this: the poorest person pays 0 extra, and the richest person pays 26000 extra, with the burden going up in a straight line in between (so the median person pays 13000 extra, and gets 13000 benefit, for a net of nothing. On this way of paying for it, the bottom half of the population gets more than they pay, so there is no cost to them. The top half of the population pays 6500 (more than they get) on average. So the total cost in this sense is exactly a quarter of what you get if you calculate it the flatfooted way in the original post (or about 750 billion). Of course, a real tax wouldn’t look like this either, and the calculation would be more complicated.
In general, neither of these are DIRECTLY relevant to how good the policy is (what is directly relevant might be things like the amount of net well-being it produces, etc.) and I don’t think there’s a very simple story to tell about its indirect relevance either, so I think it’s a mistake to focus on any of these notions of cost in arguing against a policy.
LikeLike
Thanks – I would love to see someone work the budget all the way through.
I.e.:
1) Everybody gets $13,000/year. (Including children of all ages? Including green card residents? Including illegal aliens?) There are N people, so the baseline cost is X.
2) However, we are going to eliminate the following social programs and transfers, resulting in a savings of Y.
3) In addition, we eliminate the standard deduction, the dependent exemption or deduction or whatever it is, and all itemized deductions that result in tax savings of less than $13K time family size, so we expect to see Z more in tax revenue.
4) Therefore, for static assumptions at our first pass, we think we need (X-Y-Z) in additional revenue to fund the program. You could get that money by increasing income taxes A% above existing levels, or by raising an equivalent amount by a VAT.
5) Of course, static assumptions aren’t super realistic – here’s how we think the UBI will affect trends. (5 is extra credit).
LikeLike
I don’t see how hiring a permanent live-in home care aide counts as “living independently”. I suspect that many of the people who think severely disabled people should be killed would choose death over life as a quadriplegic. Moreover, we could easily reduce that 19 dollars an hour to a more manageable figure by simply reducing the credentialing requirements for people who wish to be home care aides, similar to how we could make childcare and teaching less expensive.
It’s not immediately obvious that creating an underclass is unacceptable. Glen Weyl had a fascinating paper by the title of “The Openness-Equality Trade-Off in Global Redistribution” about how the immigration policies of the gulf monarchies (generous social safety net/basic income for citizens, indentured servitude and practically no rights for guest workers) do more per-capita to make the third world better off (through wages paid, remittances) than any first world foreign aid program. Alternatively, we could dramatically restrict immigration in concert with providing a UBI for citizens, thus preventing the existence of an underclass without paying UBI to migrants.
LikeLike
“Independent” gets used in a funny, jargony way by disability rights types like Ozy. Here’s an explanation. http://disabilityachievementcenter.org/what-is-independent-living/ “Independent living has to do with self- determination. It is having the right and the opportunity to pursue a course of action and it is having the freedom to fail –and to learn from one’s failures, just as people without disabilities do.” It explicitly doesn’t refer to being able to do things without help. I think this is bad but a load of people do it.
How horrified non-disabled eugenicists like Peter Singer are about disabilities doesn’t correlate very well with how bad they are. I’ve only ever seen Sister Y even suggest anti-depression eugenics and Down syndrome is one of the most popular targets for eugenics even though “nearly 99% of people with Down syndrome indicated that they were happy with their lives” (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3740159/) as compared to depressed people being miserable by definition. I don’t remember any studies on quadriplegics’ life satisfaction but if you want to look at anecdotal evidence, wheeliecatholic (http://wheeliecatholic.blogspot.com/search/label/quadriplegic) seems pretty happy and fulfilled. A little too busy playing games and working as a lawyer and writing poetry to worry about death.
I agree the underclass argument is weak for a different reason. Why not just restrict it to citizens? People can become citizens eventually but they can’t come and become citizens immediately and never work another day after showing up in America.
LikeLike