I think there's another possible response, the one that is, to me, the obvious and natural one: This shows UBI doesn’t work, *and that's great*! Taken to the limit, this means that transferring wealth from productive people to unproductive people makes everyone worse off. We can stop doing that and improve the lives of everyone. Oh, and we'll fix the deficit and national debt too.
OK, that'd be an extreme reaction to this study. But I think it is directionally correct, and I haven't really seen anyone make that argument.
That’s what occurred to me, too. It seems like the UBI caused people not to work, and that’s all. It didn’t seem to help the long run status of the recipients. So since it cost money, in a long run sense it’s lowering economic growth and hurting everyone. Like a moral bad. So what makes this transfer different from any other financial transfer?
> "If you tell me I am getting $1,000/month for life, then that makes me radically richer, and also radically safer. "
Lucky few! Inflation is mentioned in your article in passing, but is really another showstopper to this line of argument. If everyone gets $1,000/month for life, then cost of living will rise around $1,000/month, and leave you not radically richer, and also not radically safer.
That's only true if the 1000/month is just a new addition created by printing money (and even then, it's not _completely_ true, just mostly true). Most UBI proponents usually advocate for it being done as a replacement for existing benefits, so no extra money is being spent, which means the inflationary effect is zero.
Even if it's a new additional benefit, if it's fully paid for by new taxes, the inflationary effect is _still_ close to zero, because no new money is being created.
Now, it doesn't have to be inflationary for some portion of the benefit to be captured by groups other than the intended recipient. If the housing market remains the way it is in many markets, then landlords can raise rent by some amount (probably not the full $1k, but maybe close in some markets), and capture most of the benefit, but that's not necessarily inflationary, it just means you aren't helping the group you think you are.
Replacing existing benefits doesn't work out from either a math or political perspective. The UBI would be quite small, and people wouldn't be willing to replace existing benefits.
Fully paying for a benefit with new taxes doesn't seem very likely either.
Maybe if economic growth picks up significantly due to AI and other technology advances, along with many jobs perhaps going away, then the outlook changes. I think studying UBI now makes sense, but I see it mostly as research on potential future scenarios rather than something we could implement with our current economic constraints.
I won't argue about the political reality. I've gradually come to the realization that doing _anything_ real in today's political climate is pretty impossible, but I don't understand the "math" comment. If the money is small that definitionally means that the benefits they are receiving currently are small, unless you are arguing that the government is somehow able to get large multiplicative value out of providing non-monetary things relative to just the cash, to which I am pretty skeptical.
> For every one dollar received, total household income excluding the transfers fell by ‘at least 21 cents,’ and total individual income fell by at least 12 cents or $1,500/year.
The fall in transfer income of $4,100/year matters a lot if you’re Giving People Money privately. That wipes out a third of gains. For government it is presumably a wash.
I don’t understand. I thought the fall was non-transfer income. How is it a wash for government?
How much was the increase in "spending on others" in dollar terms? That may tie into the "social or cultural restrictions preventing them from saving".
I thought of the same thing. Poor people often survive through social networks - when one person succeeds they share with others, and expect the same in turn. This seems rational for people who are near the line where they can't afford to live at some points in their life, and have few to no alternatives. It's a type of social insurance.
I'd love to know more about how this affected parents with children and fertility - is there a summary of that part of the data anywhere?
Also, a friend had an excellent question that the study may not answer: if overall productivity didn't change or went down, did it massively increase productivity or entrepreneurial efforts in a small number? Paying UBI to find and polish the diamonds may be a net benefit.
I guess I'm in group three but my take was more along the lines of "people working less and spending more on leisure time is the real goal, yeah?" Especially inasmuch as it was being funded by an AI mogul, we're trying to figure out if giving out slices of AI employment productivity to the people displaced by it can be a humane path to downscaling a human workforce without destroying their quality of life, right??
That seemed to be Zvi's conclusion - UBI works for its original purpose (a backup when the economy no longer has jobs for people) but not for the various other goals that have been added to it.
Politically, UBI can only work if those other goals are tangible and effective. UBI for post-scarcity isn't needed now, so even if the future requires it we cannot pass it now, given those results.
Somewhat hilarious, the study was funded by Altman, and then they published it behind a $5 paywall. I guess it's not just us naive academics and incompetent governments who get screwed over by the publishing industry.
I feel that for a number of key metrics, studies like this are useless. As mentioned, this is a 'lucky few basic income' that masks the impact of a truly universal basic income (or really, even just a basic income for all low income people).
If all low-income people are given BI, we should expect that low income people as a group will have more cash and work less on average. This means a lot more dollars chasing fewer goods and services, resulting in the main impact of BI being price inflation. And the money for this will either have to come from higher taxes, lower spending on other stuff, 'money printing', debt or a combination of all four, which will have negative impacts too.
I can't comprehend how people can think "It's a good thing if people work less, actually". It feels like a caricature of economic illiterate liberalism, but literally what do these people think the money is supposed to be spent on? It's as if they think that 'working' isn't actually productive and is simply a hoop to jump through to morally entitle you to income. But if less stuff is getting produced, the primary people impacted by this are people who are most sensitive to cost of living increases.
It's like people demanding more lockdowns during covid, who claimed that it was a grave injustice to expect workers to risk getting covid to go to work and that the only reason for this expectation was so that corporate profits weren't hurt. As if people can exist on dollars per se, even if there's nobody producing goods and services to be bought with those dollars.
You might think that the marginal utility of an extra hour of work to society is less than the utility of not working for that time would be to the worker, and they're only choosing to work because of inequitably distributed resources. In that case, redistributing wealth through UBI could be increasing total societal utility even though it's decreasing total economic production. (It might still increase household and non-market production.)
Sure, and the study seems to support that idea - they worked less because they were under less pressure to work. But if all of society does that, then fewer Things are produced. Whatever it is that these people would have been making before. It could be widgets at a factory, or lawns mowed, or baby formula (to cite a fairly recent shortage that caused lots of problems). This has two major effects: less total consumption and/or more expensive consumption. So by giving people money, we run the real risk that they end up poorer in actual material conditions. In fact, everyone may end up worse off as a whole, because the same number of people are sharing fewer things. If productivity stayed the same, the best we could hope for over all of society is that the average consumption was unchanged (productivity didn't go up, so there are no new things to consume). This may still be desirable if we want to redistribute - people at the bottom don't pay into the system but still get a benefit. If total productivity goes down, even those at the bottom may still feel worse off even after the transfer.
How many hours would you ideally have people work in order to make more things? We could all work 100 hour weeks and in theory have more than twice as many things? Or do we have the balance perfect now?
It would be silly to think that we have the balance perfect. My point, and I believe Brett's point, is that we should be aware of the consequences of our choices when making big policy decisions.
More specifically with this case, we should be aware that if we all have more money, but there are no new Things, then nobody is actually helped by the program. Cash money is a piece of paper with almost no utility, so having more of them doesn't serve any actual purpose. If we all have more money, but there are *less* Things, then we are all worse off than we are now. You may think that's a good thing because you think we work too hard already, but politically it's a disaster as people are promised more financial freedom and end up with more poverty.
I'm not an economist, but that outcome doesn't seem likely to me? If prices rise because people are working less, then people will work more so they can afford the new prices, putting us back where we started (with inflation).
Meanwhile there's so many other effects on the economy that make it hard to predict how it would play out (and which weren't tested in this study):
- it eliminates the need for a minimum wage, opening up new types of employment
- it gives workers more bargaining power
- makes it easier to have children, who will grow up to contribute to the economy
- allows people to take risks in their careers
- allows people to spend time on things that they value, but which money doesn't buy
Well, even with AGI the marginal value of human labor is unlikely to be zero. If you pay the human a small enough amount per hour, there will be tasks they can do that would cost more to get an AI to do them. It just won't be enough money for the human to live on.
I'm sure there's a hundred reasons this can't work, but has serious consideration been given to the idea of the government paying for increases to minimum wage? So if minimum wage is $10, the government pays for an extra e.g. $5 per hour to make the effective minimum wage $15. This could potentially be paid for with a levy on the biggest employers in the country. In doing so, poor people get paid more *if* they continue working (to ensure that increases in wages don't lead to less production and therefore price inflation), and this is paid for by corporations without increasing the marginal cost of employing people or giving people more hours that a normal minimum wage increase would.
Hey, under those terms, I'll offer to hire you for an extra 80 hours/week at "minimum" wage if you'll hire me for an extra 80 hours/week at minimum wage. No actual work required, we'll just each pay each other $800 and then the government will give us each an extra $400.
Nice scam! Unfortunately this would not work at least where I live as there are other costs with employment put on the employer. And the risk of getting caught..
Not saying this applies to you, but for many its a weird tendency to look down on people taking welfare money, but its seen as an accomplishment to scam the government for the same money, regardless if you need extra money.
What the Op was hinting at is probably negative tax as an alternative or addition to UBI.
If you're just looking for redistribution, I guess that's not a bad plan. It forces people to work to get the benefit. Total consumption can't go up to meet the new wages, because nothing new is being produced. This will still cause inflation inasmuch as the newly higher-paid workers will consume more, unless there's a commensurate decrease in consumption at higher income brackets that pay more taxes. So, redistribution - IF - it works, or inflation if that doesn't pan out.
Re UBI will just be inflated away if universal, because actual production would fall...
I've always thought UBI was explicitly about "in a world of technologically-mediated unemployment." Presumably, when robots or AI are so good they're replacing a lot of jobs, robots are ALSO good enough to be doing a lot of the production. Then you just need to allocate some percent of that production for the UBI pop in question.
After all, how valuable *really* is it to have some moke staffing a line as a burger or gas station cashier, or restocking at a big box store, or whatever. Those are obvious "better done by robots" jobs, as soon as the robots are up to par, and that's the supermajority of jobs at this level. So working less in that world is good and expected, and that world is coming pretty soon.
Housing is the real issue then, because that's probably the last thing we'll figure out how to produce cheaply or easily, even with GPT-8 galaxy brains working on it.
Part of the problem we have now is that wages have not kept up with increasing productivity for a long time. UBI can also be seen as a reaction to that, trying to reverse some of the damage, even before the robots come for our jobs.
If we are unable to redistribute wealth better it seems unevitable that we will see more and more social unrest, destabelizing the whe whole of society. We can already see this happening it seems.
I belive the added money supply (money printing) is how modern democracy tries to adress the lack of redistribution, without adressing it, currently. And that this might cause even more problems and inflation in the future than UBI would ever do.
Well, I'll be a good Bayesian and update downwards on UBI then. No study can ever be perfect, but this was a pretty heroic effort, and flawed evidence is still evidence. There's certainly still some cases to be made - administrative streamlining is underrated, bribing people away from destructive burn-it-all-down ideologies is plausibly great ROI - but the utopian ideals from yesteryear's treatments of the subject have definitely been put paid. I do think it's encouraging that cash transfers to the deeply poor, like GiveDirectly's work, still seems efficacious...there probably is some low baseline of, uh, "societal wealth" or whatever that we'll still get productivity juice from the transfer squeeze up to that point. But it's definitely pretty low if literally giving wads of cash to the homeless doesn't change much.
(On a personal level, an extra $1k/mo post-tax would be basically an additional paycheck for me...and, honestly, it'd be a toss-up between working more and working less. Lots of trivial inconveniences I'm frugally avoiding making disappear with money, which absolutely could be, and that'd eke out up to 20% more work hours per week that otherwise disappear to inconsistent attendance and extended unpaid breaks. On the other hand, being able to afford dropping down to a 4-day week would also be wonderfully refreshing. Two days is never quite long enough for a restful weekend...)
Really appreciate the RTFP reporting as well, very glad I didn't get shitty coverage of this from elsewhere. File Drawer Effect: it's not just at the journal level...which is also a very useful thing to update on, even for one already resigned to employing Bounded Distrust on the regular. Constant Vigilance!
I'm probably in the range where the money isn't enough to change how much I work (and I'm salaried anyway, so it's hard to downgrade without changing jobs entirely). It's not so little that I would want to be frivolous with it, but it's be a slush fund for more vacations or small renovations or something like that, with a mixture of savings. Or, quite likely, my taxes go up by more than I get from the UBI so I'm slightly worse off overall.
I would suspect similar results based on how and when this was conducted, but I would phrase it as: The results were more disappointing than expected.
Choosing a selection of low income persons only in a very time limited window is not exactly what UBI even is. You could almost do this study just on lottery winners from low income households and probably see the same results. In my view an experiment like this needs to be conducted in a region across everyone living there, ideally over a longer period like 10 years.
I also think the US a particulary bad place for this study, here it would probably have a bigger impact for everyone to repair access to education and healthcare first. This would probably leave more money on the table for UBI as well. To see better reproductive results you could alternatively add payout for every child as well, but you risk ending up with "UBI-moms". (Some might not think thats a bad thing)
In general its better to tax the stuff we want less of, and incentivice what we want more of. So dont tax actual normal work, but the stuff we want less of, like polution, unhealthy foods, exessive amounts of work, exessive amounts of income, luxury items, etc etc. Perhaps we do not even need UBI for a long time, hopefully.
I strongly doubt the results related to alcohol and drugs. If you gave me an extra $1000/mo, and I used that to do more drinking or spend more on drugs, there's no chance I would admit to it. I might even feel like downplaying the amount I spent on these things even if it didn't increase with the UBI payment.
I am a bit confused by reactions to this UBI study, and I am glad Zvi Moshowitz wrote about it with his usual clarity.
Yes, this was clearly not "UBI". Given the limitations (not a huge amount of money, for a defined and not very long time), the results are not determinative of anything.
But, but, given those limitations, the results _do_ provide evidence of the directions that the effects might go in, and my expectation would be that as we move closer to a real "UBI" that movement would be magnified.
This suggests to me to that we should continue to support UBI as a replacement for existing transfers, but mainly to reduce costs and unfairness within the current system as well as to mitigate the ridiculously high EMTRs that people can face today due to benefit tapering/ceilings.
Whether as part of such a remake of the transfer system we reallocate spending towards parents or we somewhat increase it in order to be more generous to parents, as Zvi suggests for obvious reasons, is a separate question.
But based on these findings it does appear that, even assuming that UBI was taxable income, in America the change will be fiscally negative whereas in Europe with VAT rates at generally around 20% and welfare administration costs much higher, it might be more of a wash if not even slightly net positive (probably depends on how hefty payroll taxes on lower earners are, but possible).
"I see essentially four responses."
I think there's another possible response, the one that is, to me, the obvious and natural one: This shows UBI doesn’t work, *and that's great*! Taken to the limit, this means that transferring wealth from productive people to unproductive people makes everyone worse off. We can stop doing that and improve the lives of everyone. Oh, and we'll fix the deficit and national debt too.
OK, that'd be an extreme reaction to this study. But I think it is directionally correct, and I haven't really seen anyone make that argument.
That’s what occurred to me, too. It seems like the UBI caused people not to work, and that’s all. It didn’t seem to help the long run status of the recipients. So since it cost money, in a long run sense it’s lowering economic growth and hurting everyone. Like a moral bad. So what makes this transfer different from any other financial transfer?
> "If you tell me I am getting $1,000/month for life, then that makes me radically richer, and also radically safer. "
Lucky few! Inflation is mentioned in your article in passing, but is really another showstopper to this line of argument. If everyone gets $1,000/month for life, then cost of living will rise around $1,000/month, and leave you not radically richer, and also not radically safer.
That's only true if the 1000/month is just a new addition created by printing money (and even then, it's not _completely_ true, just mostly true). Most UBI proponents usually advocate for it being done as a replacement for existing benefits, so no extra money is being spent, which means the inflationary effect is zero.
Even if it's a new additional benefit, if it's fully paid for by new taxes, the inflationary effect is _still_ close to zero, because no new money is being created.
Now, it doesn't have to be inflationary for some portion of the benefit to be captured by groups other than the intended recipient. If the housing market remains the way it is in many markets, then landlords can raise rent by some amount (probably not the full $1k, but maybe close in some markets), and capture most of the benefit, but that's not necessarily inflationary, it just means you aren't helping the group you think you are.
Replacing existing benefits doesn't work out from either a math or political perspective. The UBI would be quite small, and people wouldn't be willing to replace existing benefits.
Fully paying for a benefit with new taxes doesn't seem very likely either.
Maybe if economic growth picks up significantly due to AI and other technology advances, along with many jobs perhaps going away, then the outlook changes. I think studying UBI now makes sense, but I see it mostly as research on potential future scenarios rather than something we could implement with our current economic constraints.
I won't argue about the political reality. I've gradually come to the realization that doing _anything_ real in today's political climate is pretty impossible, but I don't understand the "math" comment. If the money is small that definitionally means that the benefits they are receiving currently are small, unless you are arguing that the government is somehow able to get large multiplicative value out of providing non-monetary things relative to just the cash, to which I am pretty skeptical.
Math is discussed in detail here: https://www.milkenreview.org/articles/the-pragmatic-case-for-a-universal-basic-income
Podcast episode for this post:
https://open.substack.com/pub/dwatvpodcast/p/on-the-ubi-paper?r=67y1h&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
> For every one dollar received, total household income excluding the transfers fell by ‘at least 21 cents,’ and total individual income fell by at least 12 cents or $1,500/year.
The fall in transfer income of $4,100/year matters a lot if you’re Giving People Money privately. That wipes out a third of gains. For government it is presumably a wash.
I don’t understand. I thought the fall was non-transfer income. How is it a wash for government?
How much was the increase in "spending on others" in dollar terms? That may tie into the "social or cultural restrictions preventing them from saving".
I thought of the same thing. Poor people often survive through social networks - when one person succeeds they share with others, and expect the same in turn. This seems rational for people who are near the line where they can't afford to live at some points in their life, and have few to no alternatives. It's a type of social insurance.
I'd love to know more about how this affected parents with children and fertility - is there a summary of that part of the data anywhere?
Also, a friend had an excellent question that the study may not answer: if overall productivity didn't change or went down, did it massively increase productivity or entrepreneurial efforts in a small number? Paying UBI to find and polish the diamonds may be a net benefit.
I guess I'm in group three but my take was more along the lines of "people working less and spending more on leisure time is the real goal, yeah?" Especially inasmuch as it was being funded by an AI mogul, we're trying to figure out if giving out slices of AI employment productivity to the people displaced by it can be a humane path to downscaling a human workforce without destroying their quality of life, right??
My more extensive description for this is in part III of: https://scpantera.substack.com/p/age-of-eye
That seemed to be Zvi's conclusion - UBI works for its original purpose (a backup when the economy no longer has jobs for people) but not for the various other goals that have been added to it.
Politically, UBI can only work if those other goals are tangible and effective. UBI for post-scarcity isn't needed now, so even if the future requires it we cannot pass it now, given those results.
Somewhat hilarious, the study was funded by Altman, and then they published it behind a $5 paywall. I guess it's not just us naive academics and incompetent governments who get screwed over by the publishing industry.
I feel that for a number of key metrics, studies like this are useless. As mentioned, this is a 'lucky few basic income' that masks the impact of a truly universal basic income (or really, even just a basic income for all low income people).
If all low-income people are given BI, we should expect that low income people as a group will have more cash and work less on average. This means a lot more dollars chasing fewer goods and services, resulting in the main impact of BI being price inflation. And the money for this will either have to come from higher taxes, lower spending on other stuff, 'money printing', debt or a combination of all four, which will have negative impacts too.
I can't comprehend how people can think "It's a good thing if people work less, actually". It feels like a caricature of economic illiterate liberalism, but literally what do these people think the money is supposed to be spent on? It's as if they think that 'working' isn't actually productive and is simply a hoop to jump through to morally entitle you to income. But if less stuff is getting produced, the primary people impacted by this are people who are most sensitive to cost of living increases.
It's like people demanding more lockdowns during covid, who claimed that it was a grave injustice to expect workers to risk getting covid to go to work and that the only reason for this expectation was so that corporate profits weren't hurt. As if people can exist on dollars per se, even if there's nobody producing goods and services to be bought with those dollars.
You might think that the marginal utility of an extra hour of work to society is less than the utility of not working for that time would be to the worker, and they're only choosing to work because of inequitably distributed resources. In that case, redistributing wealth through UBI could be increasing total societal utility even though it's decreasing total economic production. (It might still increase household and non-market production.)
Sure, and the study seems to support that idea - they worked less because they were under less pressure to work. But if all of society does that, then fewer Things are produced. Whatever it is that these people would have been making before. It could be widgets at a factory, or lawns mowed, or baby formula (to cite a fairly recent shortage that caused lots of problems). This has two major effects: less total consumption and/or more expensive consumption. So by giving people money, we run the real risk that they end up poorer in actual material conditions. In fact, everyone may end up worse off as a whole, because the same number of people are sharing fewer things. If productivity stayed the same, the best we could hope for over all of society is that the average consumption was unchanged (productivity didn't go up, so there are no new things to consume). This may still be desirable if we want to redistribute - people at the bottom don't pay into the system but still get a benefit. If total productivity goes down, even those at the bottom may still feel worse off even after the transfer.
How many hours would you ideally have people work in order to make more things? We could all work 100 hour weeks and in theory have more than twice as many things? Or do we have the balance perfect now?
It would be silly to think that we have the balance perfect. My point, and I believe Brett's point, is that we should be aware of the consequences of our choices when making big policy decisions.
More specifically with this case, we should be aware that if we all have more money, but there are no new Things, then nobody is actually helped by the program. Cash money is a piece of paper with almost no utility, so having more of them doesn't serve any actual purpose. If we all have more money, but there are *less* Things, then we are all worse off than we are now. You may think that's a good thing because you think we work too hard already, but politically it's a disaster as people are promised more financial freedom and end up with more poverty.
I'm not an economist, but that outcome doesn't seem likely to me? If prices rise because people are working less, then people will work more so they can afford the new prices, putting us back where we started (with inflation).
Meanwhile there's so many other effects on the economy that make it hard to predict how it would play out (and which weren't tested in this study):
- it eliminates the need for a minimum wage, opening up new types of employment
- it gives workers more bargaining power
- makes it easier to have children, who will grow up to contribute to the economy
- allows people to take risks in their careers
- allows people to spend time on things that they value, but which money doesn't buy
- etc
You are right when discussing a world without AGI. But “working” will no longer be necessary if the marginal value of human labor drops to zero.
Well, even with AGI the marginal value of human labor is unlikely to be zero. If you pay the human a small enough amount per hour, there will be tasks they can do that would cost more to get an AI to do them. It just won't be enough money for the human to live on.
I'm sure there's a hundred reasons this can't work, but has serious consideration been given to the idea of the government paying for increases to minimum wage? So if minimum wage is $10, the government pays for an extra e.g. $5 per hour to make the effective minimum wage $15. This could potentially be paid for with a levy on the biggest employers in the country. In doing so, poor people get paid more *if* they continue working (to ensure that increases in wages don't lead to less production and therefore price inflation), and this is paid for by corporations without increasing the marginal cost of employing people or giving people more hours that a normal minimum wage increase would.
Hey, under those terms, I'll offer to hire you for an extra 80 hours/week at "minimum" wage if you'll hire me for an extra 80 hours/week at minimum wage. No actual work required, we'll just each pay each other $800 and then the government will give us each an extra $400.
Nice scam! Unfortunately this would not work at least where I live as there are other costs with employment put on the employer. And the risk of getting caught..
Not saying this applies to you, but for many its a weird tendency to look down on people taking welfare money, but its seen as an accomplishment to scam the government for the same money, regardless if you need extra money.
What the Op was hinting at is probably negative tax as an alternative or addition to UBI.
Sure, the point I was trying to make through my example is that the incentives are a bit strange.
Yeah, the negative tax angle is probably more palatable for more folks, and a bit more easy to understand.
If you're just looking for redistribution, I guess that's not a bad plan. It forces people to work to get the benefit. Total consumption can't go up to meet the new wages, because nothing new is being produced. This will still cause inflation inasmuch as the newly higher-paid workers will consume more, unless there's a commensurate decrease in consumption at higher income brackets that pay more taxes. So, redistribution - IF - it works, or inflation if that doesn't pan out.
Isn't that sorta the Earned Income Tax Credit/EITC?
Re UBI will just be inflated away if universal, because actual production would fall...
I've always thought UBI was explicitly about "in a world of technologically-mediated unemployment." Presumably, when robots or AI are so good they're replacing a lot of jobs, robots are ALSO good enough to be doing a lot of the production. Then you just need to allocate some percent of that production for the UBI pop in question.
After all, how valuable *really* is it to have some moke staffing a line as a burger or gas station cashier, or restocking at a big box store, or whatever. Those are obvious "better done by robots" jobs, as soon as the robots are up to par, and that's the supermajority of jobs at this level. So working less in that world is good and expected, and that world is coming pretty soon.
Housing is the real issue then, because that's probably the last thing we'll figure out how to produce cheaply or easily, even with GPT-8 galaxy brains working on it.
Part of the problem we have now is that wages have not kept up with increasing productivity for a long time. UBI can also be seen as a reaction to that, trying to reverse some of the damage, even before the robots come for our jobs.
If we are unable to redistribute wealth better it seems unevitable that we will see more and more social unrest, destabelizing the whe whole of society. We can already see this happening it seems.
I belive the added money supply (money printing) is how modern democracy tries to adress the lack of redistribution, without adressing it, currently. And that this might cause even more problems and inflation in the future than UBI would ever do.
Well, I'll be a good Bayesian and update downwards on UBI then. No study can ever be perfect, but this was a pretty heroic effort, and flawed evidence is still evidence. There's certainly still some cases to be made - administrative streamlining is underrated, bribing people away from destructive burn-it-all-down ideologies is plausibly great ROI - but the utopian ideals from yesteryear's treatments of the subject have definitely been put paid. I do think it's encouraging that cash transfers to the deeply poor, like GiveDirectly's work, still seems efficacious...there probably is some low baseline of, uh, "societal wealth" or whatever that we'll still get productivity juice from the transfer squeeze up to that point. But it's definitely pretty low if literally giving wads of cash to the homeless doesn't change much.
(On a personal level, an extra $1k/mo post-tax would be basically an additional paycheck for me...and, honestly, it'd be a toss-up between working more and working less. Lots of trivial inconveniences I'm frugally avoiding making disappear with money, which absolutely could be, and that'd eke out up to 20% more work hours per week that otherwise disappear to inconsistent attendance and extended unpaid breaks. On the other hand, being able to afford dropping down to a 4-day week would also be wonderfully refreshing. Two days is never quite long enough for a restful weekend...)
Really appreciate the RTFP reporting as well, very glad I didn't get shitty coverage of this from elsewhere. File Drawer Effect: it's not just at the journal level...which is also a very useful thing to update on, even for one already resigned to employing Bounded Distrust on the regular. Constant Vigilance!
I'm probably in the range where the money isn't enough to change how much I work (and I'm salaried anyway, so it's hard to downgrade without changing jobs entirely). It's not so little that I would want to be frivolous with it, but it's be a slush fund for more vacations or small renovations or something like that, with a mixture of savings. Or, quite likely, my taxes go up by more than I get from the UBI so I'm slightly worse off overall.
I would suspect similar results based on how and when this was conducted, but I would phrase it as: The results were more disappointing than expected.
Choosing a selection of low income persons only in a very time limited window is not exactly what UBI even is. You could almost do this study just on lottery winners from low income households and probably see the same results. In my view an experiment like this needs to be conducted in a region across everyone living there, ideally over a longer period like 10 years.
I also think the US a particulary bad place for this study, here it would probably have a bigger impact for everyone to repair access to education and healthcare first. This would probably leave more money on the table for UBI as well. To see better reproductive results you could alternatively add payout for every child as well, but you risk ending up with "UBI-moms". (Some might not think thats a bad thing)
In general its better to tax the stuff we want less of, and incentivice what we want more of. So dont tax actual normal work, but the stuff we want less of, like polution, unhealthy foods, exessive amounts of work, exessive amounts of income, luxury items, etc etc. Perhaps we do not even need UBI for a long time, hopefully.
I strongly doubt the results related to alcohol and drugs. If you gave me an extra $1000/mo, and I used that to do more drinking or spend more on drugs, there's no chance I would admit to it. I might even feel like downplaying the amount I spent on these things even if it didn't increase with the UBI payment.
Then at least for your health, its a good thing there is no UBI :)
I am a bit confused by reactions to this UBI study, and I am glad Zvi Moshowitz wrote about it with his usual clarity.
Yes, this was clearly not "UBI". Given the limitations (not a huge amount of money, for a defined and not very long time), the results are not determinative of anything.
But, but, given those limitations, the results _do_ provide evidence of the directions that the effects might go in, and my expectation would be that as we move closer to a real "UBI" that movement would be magnified.
This suggests to me to that we should continue to support UBI as a replacement for existing transfers, but mainly to reduce costs and unfairness within the current system as well as to mitigate the ridiculously high EMTRs that people can face today due to benefit tapering/ceilings.
Whether as part of such a remake of the transfer system we reallocate spending towards parents or we somewhat increase it in order to be more generous to parents, as Zvi suggests for obvious reasons, is a separate question.
But based on these findings it does appear that, even assuming that UBI was taxable income, in America the change will be fiscally negative whereas in Europe with VAT rates at generally around 20% and welfare administration costs much higher, it might be more of a wash if not even slightly net positive (probably depends on how hefty payroll taxes on lower earners are, but possible).