Philosopher Matt Zwolinski, co-author of Universal Basic Income: What Everyone Needs to Know, was a core member of the old Bleeding Heart Libertarians blog, which shut down in 2020. Now’s he’s singledly-handed revived the BHL brand on his new Bleeding Heart Libertarian substack. Matt recently published a critique of my response to Chris Freiman on the UBI, and he’s has kindly agreed to let me cross-post his critique here. Enjoy!
My friends Bryan Caplan and Chris Freiman have been debating the merits of a Universal Basic Income. Chris supports the idea. Bryan does not.
In his latest response to Chris, Bryan makes a few different arguments, all of which focus on the likely costs and benefits of a UBI. Bryan’s general charge is that Chris isn’t looking sufficiently hard at the empirical evidence. A UBI would be more costly and less beneficial, Bryan argues, than the system we currently have.
One of the things I tried very hard to do in my recent book with Miranda Perry-Fleischer - Universal Basic Income: What Everyone Needs to Know - is take these empirical issues seriously. Philosophy has a role to play in arguing for or against a UBI. But a lot is going to depend on how the numbers add up, or fail to add up. And a lot of that is dependent on detailed questions of institutional design - questions which, too often, get short shrift in debates over the UBI.
In this post, I want to weigh in on two of the specific arguments that Bryan makes. I’ll include an excerpt from my book that responds directly to one of them. But if you’re looking for more breadth and depth about the pros and cons of a UBI - I’d highly recommend checking out UBI:WENTK! (Also be sure to check out Miranda’s paper with Daniel Hemel, which dives even deeper into the wonky details of policy design)
Parental Irresponsibility?
For instance, Bryan argues that cash transfers to parents of young children would be a bad idea, since the fact that such parents require taxpayer assistance is evidence of their irresponsibility, and “they may spend it on alcohol.”
But this is a case where Bryan is the one who is not paying enough attention to empirical evidence. We actually know a good deal about how cash transfers affect children. In particular, evidence from the 2021 temporary expansion of the Child Tax Credit shows that cash transfers led childhood poverty to fall to their lowest level on record: 5.2%. When that expansion ended in 2022, child poverty more than doubled almost immediately, rising to 12.4%.
None of this should be terribly surprising. According to a report from the Niskanen Center, the US spends less cash transfers for children than any other country in the OECD. And, unsurprisingly, the US also has the highest post-tax, post-transfer child poverty rates of any other country in the developed world.
So, yes, some parents will probably spend some of the money they receive from cash transfers on alcohol, or worse. But we don’t need to speculate about what the overall results of increased cash transfers would be. We know. Fewer children living in poverty, along with all the short-term and long-term benefits that relief brings with it.
Too Expensive?
Bryan’s other main worry is the cost of the UBI. This, I grant, is a reasonable concern. But Bryan exaggerates the problem grossly in his closing challenge to Chris:
Forget every other issue. Simple cost estimates for a UBI are simply astronomical. $10,000 a person times 330M Americans is $3.3 trillion. That’s more than double what the U.S. will spend on Social Security in 2023. I have immense respect for your intellect, Chris. But seriously, how can you get around numbers like that?!
Of course 3.3 trillion is an insane amount of money. But no one - not Andrew Yang, not Charles Murray, not Chris Freiman - no UBI advocate proposes simply giving $10,000 to every American and leaving it at that. This is simply a straw man.
It’s not always obvious until you read the fine print, by every serious advocate of a UBI proposes means-testing of one sort or another. Either they means-test on the front-end, as Milton Friedman proposed with his idea of the Negative Income Tax. Or they means-test on the back-end. In the former case, you only cut checks to people whose income falls below a certain threshold. In the latter, you cut checks to everyone, but then tax some or all of it back from higher earners. Either way, not everyone winds up a net gainer from the program. Lower-income people do, higher-income people don’t.
So how much would a UBI end up costing? Well, it depends on how big the UBI is, and who is eligible to receive it. As Miranda Perry Fleischer and I write in our book,
A monthly UBI of $500 or $1,000 that included seniors but not children and teens would have a respective gross cost of $1.514 trillion or $3.029 trillion; including children but excluding seniors age 65 years and over would have a respective gross cost of $1.663 trillion or $3.326 trillion. A UBI that excluded both groups would cost even less, $1.168 trillion and $2.335 trillion respectively…
These numbers sound huge. But they are gross numbers, and represent the true price tag only if a UBI was layered entirely on top of existing welfare programs, without replacing a single one (the UBI-plus approach). As we just saw, some proponents advocate for this approach as a theoretical matter. Yet a $500- or $1,000- per- month UBI of this type would require trillions of dollars of new funding, almost certainly in the form of new taxes. As a practical matter, it is unlikely that voters in most jurisdictions would approve such a plan…
[T]he cost of a $500 per- person per- month UBI that replaced most current welfare programs in the United States would be roughly 7% of GDP. Government spending in the United States is currently around 38% of GDP, compared to 49% of GDP in Norway and 50% in Sweden. A $500- per- month UBI would keep the ratio of US government spending to GDP below Nordic levels, while a $1,000- per- month UBI would vault us ahead of Denmark (55%) and just behind Finland and France, both of whom clock in at 57%.
Note that the cost estimates above assume zero means-testing, either on the front-end or back-end. The net costs of either a Negative Income Tax or a UBI with a phaseout/surtax would thus be considerably lower.
The cost of a UBI is a real concern. And there are some plans out there, especially those coming from the political left which seek to graft a UBI on top of existing welfare programs, that would be outrageously expensive. But to get a sense of how much a UBI would cost, and how bearable such a cost would be, we need to look closely at the full variety of options on the table, in a way that goes beyond back-of-the-envelope arithmetic.
Having done that, I believe that a UBI of $500 per month would make an important difference in the lives of millions of Americans; that it would do so without breaking the bank; and that it would do so without causing a mass exodus from the labor market. A poorly designed UBI could be a disaster. But a well-designed one could be significantly better than the status quo.
[How we would actually go about getting a well-designed UBI through the political process is a notoriously difficult question, and one that I’m sure Bryan will press me on! But for now I leave that conversation to another day.]
Subscribe to Bet On It
Caplan and Candor
Zwolinski says that parents do not waste cash benefits for kids based on data showing that cash benefits reduce child poverty. But how is child poverty measured--based on the tangible things kids are receiving (food, clothes, etc.) or based on the reported income of their families? I don't know, but if it's the latter, then the data prove nothing about whether parents waste cash benefits for kids on things that don't benefit them.
At least according to this information, child poverty is measured based on family income: https://www.childtrends.org/publications/lessons-from-a-historic-decline-in-child-poverty-how-poverty-is-measured-in-the-united-states
If that's correct, then the writer is extraordinarily foolish or dishonest.
"No serious UBI proponent suggests giving money to everyone without means-testing"
Feels like motte-and-bailey. I haven't seen means-testing to be a common part of UBI proposals and it obviously violates the name itself.