In 1910, one out of 20 of the American workforce was on the railways. In the late 1940s, 350,000 manual telephone operators worked for AT&T alone. In the 1950s, elevator operators by the hundreds of thousands lost their jobs to passengers pushing buttons. Typists have vanished from offices. But if blacksmiths unemployed by cars or TV repairmen unemployed by printed circuits never got another job, unemployment would not be 5 percent, or 10 percent in a bad year. It would be 50 percent and climbing. Each month in the United States—a place with about 160 million civilian jobs—1.7 million of them vanish. Every 30 days, in a perfectly normal manifestation of creative destruction, over 1 percent of the jobs go the way of the parlor maids of 1910. Not because people quit. The positions are no longer available. The companies go out of business, or get merged or downsized, or just decide the extra salesperson on the floor of the big-box store isn’t worth the costs of employment.
This is also plays to a contention that I see quite a lot on social media and other places that a “solution” to technology-caused unemployment and poverty is a universal basic income, paid for by taxing robots/capital. The whole notion frankly strikes me as intellectual and economics snake-oil: a tax on robots is a tax on capital, and reducing returns from investing in capital will, in my view, reduce long-term productivity gains and hence rewards to labour that we see every time that productivity has improved. After all, by “capital”, we also must consider human capital (skills, aptitudes, moral character, even) and how is one to distinguish that from simple “labour”? (This is, by the way, a killer argument against the Labour Theory of Value that underpins the rickety structure of Marxist economics.)
Here is an article at Econlog casting doubt on the “robots taking our jobs” theory, while pointing to a debate on the subject worth looking at.
I had a brief comment about this on Samizdata before, on May 23.
I have several problems with UBI, along the lines suggested by a writer at Catallaxy Files here. Bryan Caplan has written another, in my view, strong take-down of the idea. Yes, I know that a variation on this is a negative income tax, an idea embraced by no less a figure than Milton Friedman.
The attractions in superficial form of UBI are obvious, not least its apparent simplicity, and the idea that one could cut through the current morass of state entitlements/subsidies etc and even bolster support for a free enterprise system if everyone gets at least some sort of payment. For me, however, an issue is more moral – the idea that one is entitled to a handout by simply being a living, breathing creature – and economic – the potentially deadly impact on incentives and character.
I am planning to give a couple of talks about this subject in London, with one definite commitment being at Brian Micklethwait’s place at the end of September, and other possible talk in August. Details are forthcoming. I am generally, I think, against the idea, but I am happy to hear and read any really strong cases for it if people want to suggest them in the comments.
While it is clear that automation has reduced employment in many areas, and not just low-skilled jobs, as the loss of entry level jobs in engineering attests, it should be noted that the ongoing reduction in real income among the working class is also driven by women entering the workforce (nearly all unskilled), immigration (both legal and illegal), off-shoring manufacturing and jobs, and free trade.
You can argue, and I will accept it, that these policies have increased real GDP, but you also have to admit that they have immiserated the working class, and that all the gains have accrued to the Ruling Class.
Citizen’s Income is an improvement on what we have. Complaining that one should not be entitled to a handout simply for existing is missing the point: one is already entitled to a handout simply for existing, but it’s a handout that gets withdrawn as you earn money, causing effective high marginal tax rates and decreasing the incentive to work. A Citizen’s Income/Negative Income Tax (they are precisely equivalent) avoids this problem.
How it is funded is a separate issue. You’re right that taxing robots is a tax on capital. But that’s irrelevant. It could be funded from existing taxes on income, capital gains, whathaveyou. It would be largely tax-revenue neutral anyway, because it would *replace* existing benefits. People who currently get benefits would get the same sum. People who currently don’t would get it, but would be taxed slightly higher towards the top end, so it would be revenue neutral. If a Citizen’s Income is funded from income tax, then it’s essentially reducing the high marginal tax rates on the poor, increasing their incentives to work, in exchange for a slight increase on marginal tax rates for the rich, slightly decreasing their incentive to work.
bob sykes,
Not sure either of your terms of reference there are helpful – unskilled work and real GDP seem to be the two poles you are opposing – as neither is an ideal for anyone other than a statist.
That jobs that require no skills (which don’t really exist – every job requires skills to do it well at least) are disappearing (because it is cheaper to get a machine to do them) and that this is starting to spread to jobs which require skills but only basic ones is an issue if you have no skills or only basic ones. But humans are adaptable – and not all low skill jobs are disappearing (I hadn’t noticed hair dressing was threatened by robots for example), so the disappearance of a set of jobs (which is gradual anyway) is only a problem if we regard the people holding those jobs as not capable of moving to other jobs. Which is the same basic process as happened after say the loss of the ‘unskilled’ agricultural jobs, which provided the manpower to fuel the industrial revoution. So it is only a major problem if you wish to for some reason keep people defined in a certain way, as ‘working class’, rather than admit that we need to move on (after all, in European history we have clearly lost the slave class, the agricultural class and the servant class over the last few hundred years – why losing another class would be a problem is beyond me).
And real GDP is a poor measure, as it could be one man making 90% of the money – as a measure GDP or even GDP per capita only tell you about the economy, not the individual.
We already have UBI for many middle class people.
The vast increase in government bureaucracy, the state education system, regulatory bodies, NGOs and the pseudo charity sector has created huge numbers of non-jobs to keep these people in the style to which they think they are entitled to.
Automation will mean that a full UBI system is needed or non-jobs will have to be found for millions of modestly skilled workers who will become an electoral problem if they are abandoned to the market.
Callous treatment of the unemployed only works in a democracy if most people can find decent jobs and can ignore the plight of the poor.
That has always been the situation so far, but things are changing fast, with real unemployment rates much higher than the official figures, and with underemployment and low pay becoming much more common.
And exchanges it for a different problem of exquisite political toxicity: the amount paid will continually rise, that is self evident. And on the basis you get less of what you tax and more of what you subsidise, more indolence appears to be a feature not a bug of such schemes.
Many years ago when I were a lad, I wrote an essay at school in which I predicted that with machines taking everybody’s jobs we would eventually end up with massive unemployment and economic collapse. Even before I got the inevitable lowest grade ever I had worked out that I was talking nonsense. So, I was somewhat surprised when I got an “A”.
As a member of Samizdata’s working class I can say that by and large the things that work for me are run by the free market and the things that don’t work for me e.g. housing and transport are either run or heavily controlled by the state.
I think the a primary argument against UBI is that it is such a horrendous waste. All these capable people who could be doing useful things to make their lives better, and the lives of everyone in society better, is wasted because they are incentivized to stay home and not work? It is the most appalling waste. And to make it worse, taxing the very capital that could fund their contributions to make it happen? Insanity.
As to @Bob Sykes comment about the poor being immiserated to the benefit of the rich I would have to cite the indomitable and sorely missed Maggie:
https://www.youtube.com/watch?v=pdR7WW3XR9c
The problem we have today is different. Whereas before those elevator operators would have gone on to do something useful and productive, nowadays most jobs being created are in sprawling bureaucracies linked either directly (public sector) or indirectly (private sector) to ever-increasing regulations and the administration of government. Corporations appear to be growing ever-larger, but with white-collar workers in process-driven functions which didn’t exist a decade or two ago, and any cuts fall on the blue-collars manning the production lines or providing the actual service. I am not sure there is much to cheer in seeing 100,000 jobs lost in manufacturing and mining while 100,000 jobs are created in various government offices or diversity and compliance departments in major corporations.
There seems to be a general tendency to consider “concepts,” such as Universal Basic Income, principally from (and sometimes limited to) perspectives of “economic analysis.”
Perhaps an equally valuable perspective may be had from considering the comparisons of “universality” with “individuality.”
For clarity: we are addressing individuality, which is comprised of the distinct characteristics of an individual, as distinct from individualism
(or individualistic),
which is the expression of those characteristics in a social context.
In universality, when it appears, there may be seen a result of universalism, which, in turn, is probably always derived from “consensus” in the social context which overrides individuality.
So we might look upon some proposed “universal” social benefit as a form of consensus collectivism that displaces (to some degree at least; and probably substantially) individuality.
The great human conflicts from classical times onward and certainly over the past 200 years have given us examples of the effects of attempts to achieve universality.
As an avid reader of McCloskey, including the more recent commentaries on libertarianism in addition to the Bourgeois
series, I noted slight consideration of the role of individualityin both the location and development of the Industrial Revolution and the differentiation of the societies of northwestern Europe. Some of what has been missed can be found in the scholarship of Alan MacFarlane and the essays of Michael Oakeshott.
Perhaps we should take far more care in considerations of “universal” arrangements, whether they be healthcare, education, income, or what have you, in observing the nature of collective consensus involved and the detriments to individuality.
UBI does not, and cannot, address the very real problems Tim Newman presents. Genuine and problematic those market distortions may be, but they are a non sequitur regarding the pros/cons of a UBI.
My primary objection to the UBI is that it requires taking more money out of the economy than will be distributed in the form of UBI benefits. This is inherent in the political process, it is unavoidable. The transfer costs would seem to be sufficient to sink the notion because no matter what figure we agree people should receive as a UBI, more money will have to be collected to fund all the new jobs involved in tracking, managing, overseeing, and distributing the UBI. It will require the build-out of a vast new bureaucracy, and those don’t come cheap. The ‘free rider’ problem crushes the horse the idea rode in on.
Some sort of UBI will become an unaffordable necessity because it seems very unlikely that a genuine robotic/AI transformation wouldn’t displace a huge amount of people from their jobs. I don’t know how far we are from such a transformation (I suspect quite some way) but it would massively disrupt the workplace and thereafter the economy – what’s the point of making stuff if no one can afford it because they’re nearly all out of a job? The economy can’t afford to pay a genuine UBI and there’s the moral issue of worklessness as highlighted above though perhaps people would eventually adapt to carve useful achievements elsewhere. If the transformation takes place quickly some sort of payment will be politically and morally required else most of the working and middle classes will be condemned to oblivion and a violent re-adjustment will be likely to follow.
And what makes you think new ways of working will not emerge as they always have?
Tim Newman,
You are confusing two separate issues there – the loss of jobs to automation (a process that has been going on since the invention of the hoe (note the e…)) and the allocation of resources within the economy. Regardless of how many bureaucrats were hired, from 0% to 100% of the population, automation will still remove jobs (even bureaucrats innovate remember – some of us like a bit more efficiency in our systems. So automation may be moving people into less ideal jobs (at least in our view – note a certain amount of bureaucracy is necessary in any inefficient system, until it is automated…), but that is simply because government is large and regulation imposes costs. If we dealt with that (difficult) problem, then resouces would be freed up to be allocated elsewhere.
I think there is an all else being equal fallacy on display amongst the universally pesimistic advocates of UBI here (hint – if your idea is only promoted pessimistically, it is a very bad idea).
The argument appears to be that sooner or later machines will replace all work people can do. Fair enough – I can theoretically see the limit of what machines can do overtaking the ability of humans to create new jobs. But this is forgetting something important – jobs in themselves are a cost on both the employers (they have to pay for them) and on those who have them (I have better – or worse – things to do with my day). So in effect by removing jobs we have removed a huge amount of costs – in effect, we have removed labour from the economy, and capital alone is producing the wealth.
And then think of what wealth – not only the wealth produced by all capital and labour today, but also the potential wealth that all capital and labour today might produce but cannot due to constraints (labour and capital is limited): if no jobs are available, the full potential use of all labour and capital must have been reached. And then we also have the fact that the limit of what is possible for labour to do will have advanced by the time that machines catch it up – new jobs using freed up labour and capital will have appeared and then disappear again – so we will not simply have the wealth that is potentially available now, but the wealth potentially available at a point in the future where labour has repeatedly built upon the fact that it is being freed up again and again to produce more wealth, right up to the point the machines catch up (the earliest prediction for this I have realistically seen is 2050, but considering a conversation I had earlier today with an IT guy about why a form on a webpage could not do a simple IF… THEN… I have my doubts on that…).
And then realise that the machines are not going to sit still, since one of the functions that they will be better at than humans will have to be innovating to get better returns on capital, as otherwise humans would still have jobs, so the wealth will continue to accrue.
So there will be a lot more wealth around by the time humans are made obselete (assuming this happens). So we are not dealing with a scarcity problem, and could well be dealing with an Iain M. Banks Culture situation (in effect everything is free because of the ability of machines to make things cheaply). So why would anyone think that the solution to this problem would be to insert the most wasteful element of the economy, government, into the equation and make them responsible for passing out wealth – in effect giving control of individuals’ fates to people who are not them. Whilst there would need to be some way to ensure the wealth is shared, that seems to happen reasonably naturally in a market/capitalist system: note how much better off the poor are now than at any time in the past – this is not the state in action since it is also the case in countries with no welfare system. So why would we go for a system that would allow government a role, when the other option is to have everyone free and unbeholden to the will of politicians, majorities or interest groups?
I’m sure there are objections to this, and I’d be interested to hear them, but remember the basic point here is that if the situation comes about which causes normally sane people around here to state UBI will be required, we will all be incredibly wealthy anyway. Thinking that UBI is the answer is thinking like a socialist – thinking you can make one change to society to solve problems, without considering the fact other changes happen anyway; thinking that somehow government is the best solution to a problem simply because it provides a superficially simple solution; and thinking of humans as having to be dependent, not independent. It is pessism and the triumph of the collectivists written large.
It kinda reminds of a Carry On movie. Sid James is malingering in hospital and his Mrs visits him and complains he is a scrounger. He responds that he can’t get work. She replies, “Well, that’s what happens if you call yourself a Hansom Cab lamp-fitter”.
I still don’t understand how we can discuss how we might someday have a UBI in what’s already a welfare/entitlement state.
If I fail to make enough money to support myself, the state will do so. Not in luxury, maybe, but I’ll be fed, I’ll have housing, and I’ll likely have some spending money.
When I quit working due to age, I’ll be (in the US) on Social Security.
Are we just talking about a more luxurious benefit? Because, so far as I can see, we have UBI now.
A few random thoughts (in no particular order) on this post:
1) Most of what passes for “wisdom” from mainstream economists is “intellectual snake oil”. Four+ generations of university-trained economists, all indoctrinated into Keynesian pseudo-economics, could have no other result. The exceptions are few and far between (McClosky being one of that rare breed). And if that “wisdom” is adopted by leftists it’s guaranteed to be pure, undiluted snake oil.
2) Milton Friedman did indeed propose a negative income tax. However, later in his career he recanted and apologized for his error.
3) Robert Heinlein actually proposed a form of UBI in his novel “For Us, the Living,” and even spent some time attempting (unsuccessfully) to describe the economic basis for it. However, it was one of his first efforts (and not a very good one at that), written at a time in his life when he was an avowed socialist. Obviously, he outgrew that adolescent phase. It was never published during his lifetime (because he came to realize that it was utter crap), and his literary executors did his memory no favors by permitting it to be published posthumously.
4) UBI is complete socialist nonsense: it is economically irrational, unsustainable, and provides precisely the wrong incentives for all concerned (those who receive it and those who are taxed to pay for it). And that’s without even considering the dead weight of the program costs noted by Shirley Knott. It is by no means equivalent to a welfare system; flawed as that system may be, there are still controls (and the ever-present possibility of expanding those controls) on who is entitled to receive it, on what terms, and for how long. None of that applies to UBI.
5) The entire concept that technology is creating some form of “structural unemployment” is fundamentally flawed. It is the 21st century equivalent of Malthusianism. Aside from the ability of people whose jobs are lost to technology to be trained for other work, the economy is dynamic. Some of those whose jobs are lost will start their own businesses. Not all, certainly (most probably shouldn’t try it), but some will, and they will hire others. Small business is the engine which drives most of the economy. It is where a large percentage of job creation occurs, and those businesses are not big users of technology. For them, even if theoretically technology could eliminate some jobs it won’t, because those small businesses don’t have the capital to invest in it. They will continue to employ human workers for such tasks, and they can do so and still be profitable because they simply don’t have the overhead and sunk costs of their larger competitors; they find niches in the “cracks”. The real problem is not technology destroying jobs (which is a good thing); it is government making it unnecessarily difficult to start new businesses. Get government out of the way and most of this “problem” disappears.
But can we automate politicians, lawyers and judges?
Perry: In reply, I’d say that all previous technology revolutions have unlocked a further layer of complexity that could only be done by Humans because Humans still had the brains. An AI/Robotics revolution would leave no new layer of complexity that couldn’t itself be done by AI/Robotics (this is all highly speculative obviously). I would assume, that the first true AI/intelligent robot could simply design a better copy of itself to do whatever new jobs might appear. Only highly artistic or leadership roles might be left for Humans and we know from history only a few Humans are good at those (though that could be from lack of opportunity in the past). I’m not saying that an AI revolution is on the cards but if it is then only giving the suddely unemployed population something to spend would we be likely to stave off some upleasant consequences. If the working and middle classes are no longer needed then do you feed them or ??? I’ve gone far off the original post here, sorry. If UBI is needed depends upon the next big thing, if the next big thing thinks for itself then there will be no new jobs in the sense that we’ve known them.
The problem as I see it is that as the rate of change accelerates which I think is axiomatic that it is doing, there is less and less time for people displaced from jobs by technology to retrain and move on.
So the buggy whip maker model doesn’t hold.
Intuitively I think UBI is a horrible idea. It reduces all to mere ciphers. The incentives are all wrong. It cannot but lead to disaster.
If you have time, could you expand on this? In my mind, a UBI is merely a floor – a guaranteed minimum income that will most often be replaced as one makes their own earned income. What controls are available in a welfare system that wouldn’t also apply to a UBI? I’m assuming a UBI is means-tested – it’s not going to be paid out to people earning their own income. Am I incorrect about that?
We may speak of controls on welfare systems, but in the end, (absent mental illness that keeps a person out of the system), no one in our society ever starves, or cannot find shelter. We supply that basic floor of income that keeps people alive. Where is the huge difference between our present welfare system and a UBI that you’re talking about?
The “revolution” we are seeing in AI and robotics is all about low hanging fruit. General reasoning is FAR beyond the capabilities of the math(s) used for AI and robotic control. Robots appear to be general purpose, but they are limitied by software, which isn’t. So large corporations will replace easily replaced human automatons with synthetic ones. But there are lots of jobs not so easily replaced. And as Laird points out, there is room at the bottom for new small efficient businesses in the “cracks”. Those cracks are defined by our human ability to specialize quickly, in a general way, something our automation systems will take a looooong time to figure out.
Laird,
I agree with most of your points here but for the small businesses don’t use tech schtick. Both me and my wife are self-employed (I’m an IT tech and web-designer, she’s a translator). Both are highly tech-dependent. We both need fast internet and at least a couple of computers each. OK, we don’t directly hire but we do indirectly create jobs by enabling things.
No one ever laments being born into ‘comfortable’ circumstances, and no one ever suggests that those provided with ‘a competence’ are doomed to live wastrel lives. The moral hazard some associate with the UBI, if it were real, would already have demonstrated itself in the lives of those born to financial independence.
As to the practicability of the UBI, I have no opinion; but the ‘moral hazard’ argument against it strikes me as ill-founded.
Myno, I’m not sure that better AI is FAR beyond where we are now – perhaps only one or two breakthroughs and who knows when some brainbox will crack those. I’m sure reliable steam power once seemed beyond “current capability” but it arrived. An AI doesn’t need to be the best at everything to do a seemingly complicated job, most people’s jobs are specialist to some degree – I’ve found that the hardest thing about quantifying most people’s work-process is getting them to tell me what they actually do (as opposed to what they’re believed to do by their organisation) and when cost-effective I’ve seen complicated job tasks replaced with conventional SW and custom algorithms – so even basic machine learning may be closer to having a larger impact than we imagine. I doubt I’ll still be working when AI becomes “good-enough” but I think that it will come eventually and then some sort of universal income may be politically required – at least for the transition to whatever comes next. Time will tell.
Spence, there are many different approaches to AI. The old “expert systems” generally failed due to lack of real world knowledge. They were so specialized that they did not justify their complexity. That effort was largely supplanted by the sort of specialized software of which you speak. The recent AI advancements of which everyone is in a tizzy are in a corner of mathematics that does not attempt to deal with general knowledge. The fancy new AI systems are all based on pattern recognition, which as my old grad school prof said, “…is everything or nothing, depending on how you look at it.” It’s everything to anything that can be reduced to pattern recognition, but that neatly defines the low hanging fruit of which I spoke. General knowledge, and the concomitant ability to utilize it, will define a breakout in AI capability. As Dante noted, prognosticaters are frauds, consigned to the 8th ring of hell, but I still think that that breakout point is decades away. When it arrives, the landscape will be quite changed by the natural progress of the market, hobbled though it may be by enlightened bureaucrats. There will be a period when AI and robotics will expand their influence, but that will be followed by easy to learn interfaces that allow non-PhDs to program them, which will engender a new resurgance in small startups that attack the cracks in the large scale AI and robotic firms. Life will continue as it always has, with displacements due to technology advancements, and opportunities for everyone at all rungs on the shifting ladders of the economy.
If I may address your steam power comment, I would dare say that those of a technological bent in the days before steam would have had little trouble imagining that there could be some new way to power the watermill tech that preceded them. The advance of AI is qualitatively different. Very smart people find it difficult to imagine the character of the breakthroughs that will enable general purpose AI.
Hello Myno, thanks for your reply. I think the thing about breakthroughs is their power to surprise; I don’t mean surprise the layman as he’s often not that interested – I mean surprise the experts. They often see complexity on all sides and fail to see incremental progress here and there, breakthroughs are the peak of a massive wave that gathers out to sea and heads towards land – experts stare out from the beach seeing a gentle wave languidly approaching and failing to anticipate the drenching that they’re going to get. I design algorithms but have never had that much interest in AI, all through my career it’s seemed ridiculous and a waste of time. But in the last year or two it has seemed to spark. You’re right, AI has seemed to spark before and then gone dead as hopes for it died, but I wonder if processing power is finally getting to the point where something better may be on the way. Interesting though; do we fear it or will it lead us to somewhere better than the world we have?
Enter First Dinosaur, stage left. Looks at sky.
Enter Second Dinosaur, stage right. Looks at First Dinosaur.
Second Dinosaur : “What you lookin’ at, hun ?”
First Dinosaur : “That fireball. I think it’s a meteorite. If it hits the Earth, that’s curtains for us.”
Second Dinosaur “Oh I shouldn’t worry, hun. Those meteorites – there’s always a big fuss about them on the TV, but they always miss.”
Better. More power at your fingertips. Less time and effort expended doing the mundane. Progress! Of course hobbled by bureaucrats, perhaps disasterously so. From which new tech will break free, only to be hobbled in its turn. Life. Beats the alternative.
@ PfP:
Would it be (and where it exists, is it) a form of collectivization?
Does collectivization carry with it “moral Hazard” for individuality?
NickM, you are not “tech-dependent” in the sense of this discussion. You use a couple of microcomputers and the internet. Big deal. The tech we’re talking about is specialized machines which displace human workers: industrial robots running assembly lines; specialized fruit-pickers which can clear a field or orchard in a very short period of time without either missing or damaging the fruit; fully-automated McDonald’s restaurants with no need for counter workers; etc. Those are industry-changing tech, and are also extremely expensive. A small custom metal fabricator isn’t going to invest in top-of-the-line computerized lathes or CAM systems (basic ones, perhaps, but those still require a human to run them); a mom-and-pop diner isn’t buying a system which could run a McDonald’s. Your job isn’t likely to be lost to automation. Your wife’s probably isn’t, either. It’s certainly possible that AI could advance sufficiently to handle routine translating tasks, but only the largest companies are going to invest in it; most of her clients, having only occasional need for a translator, will find it more economic to continue to hire her. (Although she might buy that software, just as accountants all now use tax preparation software, but it’s relatively inexpensive.)
bobby b, the devil is always in the details, but what I’ve read about UBI proposals is that everyone would qualify simply because they’re breathing, and while there might (probably would) be some sort of income cap there would be no other qualification. With welfare systems, you generally have to periodically prove that you’re looking for work unless you have some sort of exemption (small children, a disability, etc.) and there may be time limits on eligibility. Those requirements can be more or less strict, and more or less rigorously enforced, but the important fact is that they exist and you have to at least go through motions of complying with them. Not so with UBI (as I understand it, anyway). And while we don’t generally permit people to starve or go without shelter, we address those problems through such things as food stamps (now SNAP cards), which tries (however unsuccessfully) to limit the expenditure to food, and rent subsidies which are paid directly to the landlord. For the most part we don’t just give people cash to spend as they like, which a UBI program would. In my mind, that is a couple of huge differences between the two systems.
Laird : specialized fruit-pickers which can clear a field or orchard in a very short period of time without either missing or damaging the fruit
If i might just pick a nit, here. Missing or damaging fruit is just a cost. You don’t have to automate a process to replicate the way humans do it, or even the final result that humans achieve. You just need to turn a profit. So if humans destroy 5% of your crop and an automated fruit picker destroys 10% of it, that is only one strike against the automated fruit picker. It gets two more swings, and if picking 90% of the crop at a tenth of the pickin’ cost in a tenth of the time makes you more profit than picking 95% of the crop by hand, more slowly and more expensively, that’s still a home run.
Automated processes and manual (or cranial) processes often, even usually, produce substitute products rather than identical products. But they still compete. This considerably widens the field for automated competition.
I know you know this, but it’s easy to gloss past this when explaining how long it’ll be before machines can do X. Machines don’t have to do X to knock X-doing humans out of a job.
It’s not obvious to me that a UBI is necessarily impossible or destructive of incentives – so long as it is attempted in the right way.
The microeconomic problem is that if you give a chap $x for doing nothing, his incentive to seek a job paying $1.25x for 40 hours hard and boring work is rather modest. Is $0.25x worth 40 hours hard and boring work ? And what if the chap’s 40 hours is only worth $1.03X ?
So the first step, I should have thought, is to make sure that the chap only gets his $x, if he puts in 45 hours hard and boring work. This does not have to be work that needs doing. Its value to society, in itself, could be zero. It is simply there to provide the chap with an incentive to do something productive and better paid, if he can find it. So there’s a certain amount of admin cost in making sure the chap does his work – digging a hole on Monday, Wednesday, Friday; filling it in on Tuesday, Thursday, Saturday. But that kind of invigilation can probably be automated quite easily 🙂
So let’s look at some different chaps :
(a) Mr Student Bum who could get a job paying $1.25x for 40 hours work. What’s he going to do ? He’ll probably take the real job.
(b) Mr Not Completely Useless who could get a job paying $1.03x for 40 hours work – he’ll probably take that
( c) Mr Not Very Useful who could get a job paying $0.88x for 50 hours work. He’ll take the 45 hours ditch digging job – unless we can think of a way to subsidise the employer to pay him a bit more
(d) Mr Totally Useless who couldn’t get any job. He’ll do the ditch digging.
Aside from making sure that those folk who take the basic income for digging and refilling ditches actually do the spadework [thanks Mr All Seeing Robot], you don’t need to “police” this scheme. Microeconomics will police it for you. The “wrong people” won’t claim benefits*.
So the microeconomic problem is really folk who are somewhat productive, but whose productivity is worth less than the basic income. Well I didn’t claim it was a panacea. But it seems to me the cries of “Impossible !” are all based on the assumption that it has to be done by giving folk money for doing nothing. Yes that’s the way the current welfare state works, but it’s not obvious that it has to be like this.
PS Note the 45 hours ditch digging has to be labour the chap really doesn’t want to do. Sitting in the warm folding paper doilies may be dull, but the working conditions are better than lots of real jobs. So the fake job needs to be hard work, no fun and grubby. So that pretty much any real job is to be preferred. This is not sadism, it’s microeconomics.
* no basic income for immigrants
Laird,
My point is I am very tech-dependent and so is my wife. Whilst that is scaleable in the sense that we don’t use tech to the extent of HSBC that is just because we ain’t as big as HSBC. It is relative. As to our obsolescence… AI is a way away from replacing Lizzy – machine translation – apart from in very limited, very specific areas is very poor and as to me. I ain’t seeing robots crawling under desks with cables in their combat pants any time soon. Yeah, eventually but eventually I’ll be dead so it matters not a jot to me. And no we don’t have or want kids.
The post is true under capitalism – as new productive jobs are created when automation gets rid of old jobs.
However, the United States has been moving away from capitalism for a long time – and many American towns and cities (far from the Credit Bubble “financial services”,and other, elite which is sustained by the Federal Reserve) are in ruins.
But the blame for the lack of real productive jobs for so many modern Americans does NOT lay with automation – it lays with the march AWAY from capitalism over many decades.
RRS beat me to it. An elevator operator or a candle maker losing their job or business are not just another statistic – they are real people with a very real problem. Their problem is not unsolvable, but their (temporary?) hardship is very real to them, as well as to others connected with them by way of business, friendship and family. In other words, these problems are individual, as all problems really are – even if and when they are shared by vast numbers of people. Libertarians tend to forget this individualistic aspect of the dynamics we refer to as ‘economy’ almost as often as socialists do. So it’s not that unfettered free markets and tech innovation, as well as zero welfare state, are some sort of panacea – they are not, because they are still an aspect of human life, which is never a panacea; it’s just that they allow for more individualistic solutions to individual problems. UBI is the direct opposite of the latter, being collectivist in its very premise and approach to an even greater degree than the current welfare systems are.
You are confusing two separate issues there – the loss of jobs to automation (a process that has been going on since the invention of the hoe (note the e…)) and the allocation of resources within the economy.
What I’m doing is pointing out that, in a discussion on the continual destruction and creation of jobs, it is worth looking at what the destroyed jobs are being replaced with.
Alisa,
Reminds me of something I was cogitating on last night whilst reading the news. Libertarians tend to promote individual choice through arguments that carefully avoid engaging with any particular individual and talk in principles. Collectivists tend to promote collective programmes through arguments that build on the experiences of individuals and then generalise from this to society as a whole. This just seems wierd and also might give collectivists a head start in appearing sympathetic, despite the fact that you cannot argue for a group from an individual as that is a form of analogy.
This might be because collectivists who I read tend to be journalists, or using fellow-travelling journalists accounts, so this slants the sample, but I think this is a real issue. A quick scan through the first page of Samizdata reveals this to be the case here – no story is hung on the experiences of an individual (other than the author – Brian’s saga of buying a computer screen is on there). Probably no biggie – Samizdata is basically for consumption of people who think in this way – but it seems to reflect a habit of thinking. Even when I have seen a libertarian respond to a collectivist story based on an individual experience, it is notably the libertarian rarely engages with the individual.
Not saying this is wrong incidentally (I never seem to engage with individual experience myself, as I know the difference between a datum and data…). But sometimes we seem more unconcerned about individuals than the collectivists…
The biggest problem libertarians run into when thinking about UBI is that they fail to account for the fact that most people don’t think like libertarians. This is a well-known political failure mode (see socialism), but everyone thinks they are immune. So let’s expand on a few points, and keep in mind who is going to be voting for these things. Not just you, but Democrats, Republicans, Tories, Labour, Greens, etc.
1: Cost. Caplan touches here, but really think about this. If we are replacing all social safety net things, it has to be large enough for a truly destitute or disabled person to live on, otherwise those people will get dragged out every election, and the amount increased. $4k is not going to cut it. Second, the largest single expense for many of these people will be health care, and you aren’t getting health care on 10k per year, so the affordability problem is even worse. You can “solve” by having universal health care and UBI, but that only moves the affordability problem.
2: Who gets it? Citizens? Children? If you don’t limit it to citizen adults, you create a lot of bad incentives for illegal immigration and for poor people to have more kids. This is one of the worst features of the current safety net, and UBI actually makes it worse.
3: The impossibility of binding future voters/governments. The political pressure on a UBI will always be positive, and will be much stronger because so many people will be a constituent of it. All elections will be between a conservative proposing a 3% increase and a liberal proposing a 10% increase. Even if the problems are solved when it is passed, there is no way to keep them solved.
4: The misreading of the hardness of the public. UBI requires that if a poor person with six kids blows their UBI on coke and whores, the voters look at the starving kids and reprise Scrooge: “Are there no poorhouses?” Maybe you think your electorate is that hard. I know mine isn’t.
If your policy doesn’t account for human nature, it is destined to failure.
PfP,
You are aware are you not of the recurring theme in some of our press around the children of aristocrats and the wealthy behaving in ways that if replicated across society would either destroy our stock of wealth (because they are not in general producing any wealth whilst destroying it) or make life one excellent party (I feel the need to insert dude here – this is probably better than carbon dating as a way of telling my age…). That is the moral hazard of UBI aside from collectivism – the lack of requirement to provide for yourself and therefore a failure to act as a responsible individual. Whilst I think ultimately that is the utopian vision, we need to have the wealth to achieve this in place, and the imposition of UBI now or in the near future not only be too soon, but would actually (by increasing the role of government) set back the development of the required level of wealth.
Or to put it another way, UBI turns the entire country into entitled and non-productive morons.
his incentive to seek a job paying $1.25x for 40 hours hard and boring work is rather modest.
I thought the point was the “hard and boring” work will eventually be done by robots?
I’d imagine that once our robot overlords take over, everything will become so cheap you wont actually need any income, or rather, just a minor income, the balance between stuff that costs something and stuff that costs to make (wages) should always exist, just that less stuff will fall into that category.
Then again, the nuclear power dream of “energy too cheap to meter” was never realized.
And there is always the rent-seeking overhead to consider, seeing as the unions are likely to block any threat by public sector robots (see train drivers for an example).
Absolutely spot on. This is one of the reasons Trump won: sure, globalisation has been great for almost everybody, but not the poor sod who lost his factory job and has no hope of getting another. The costs of globalisation were ignored, and we should acknowledge them (as a minimum) even if we know that overall globalisation has been brilliant.
Exactly so. And it is a trend we are already seeing. I was chatting with a policemen a few years ago and he told me many kinds of property crime had been on a long term downward trend, such items as music players of various kinds, radios and televisions… because they are so cheap in relative terms that they are not really worth stealing and even people with relatively low incomes can buy them second hand with less risk than nicking one.
I assumed the talk of a UBI was just the next in a long line of socialist attempts to increase the welfare state until it achieved total coverage. It would be the end of a productive white society and I’m sure those pushing it think that is a fine goal.
Watchman, that is a very astute observation – food for thought. However, my ultimate point in that longish comment was at least as much about solutions as about problems. Think of the proverbial roads and who will build them: I’m sure all of us faced the situation where we tell someone that government should get out of this or that, with the automatic response being along the lines of ‘But what is the alternative?’. And our just-as-automatic response is to draw up all sorts of schemes, some from times past, some purely imaginary (not that there is anything wrong with trying to imagine possible solutions to future problems). Been there, done that – does not work. Instead, my new current response is: ‘Give me a specific example?’ This is supposed to make people think about real situations involving real individuals with real problems, needing real solutions – not some vague generalizations, such as ‘the poor’, ‘the infrastructure’, ‘the unemployed’, etc. If they can’t think of a real-life example with which they are actually familiar first-hand, there is no point in further discussion, and they usually see it too. (And no, ‘I read about it in a last-night’s paper’ is not enough of a familiarity with a particular situation). However, if the person can show me a case involving a friend or neighbor, etc., then we can discuss it, see how that person’s problem is being solved now by the government (don’t laugh there in the back!), and compare to the ways it could be solved absent government interference.
One mechanism that could be used to administer a UBI in the U.S. would limit transaction costs by using the existing income tax system, run by the IRS. The existing personal exemption, a.k.a., standard deduction, would simply be raised to the level of the UBI, for instance, $20,000, and would also be a credit, so that the poorest people would end up getting a check from the IRS. In this way you’re not incurring much additional cost because you’re not trying to target specific people for a benefit (and incurring all the hassle involved in vetting them and checking for eligibility). Everyone, including Oprah Winfrey, gets her standard deduction raised, even though only those below a certain income will actually get a check.
One inevitable downside of this, it seems, is that it would put wind in the sails of a movement to eliminate the cash economy, i.e., to make all transactions traceable by the IRS. Any UBI system built around reported income would naturally incentivize people to seek more unreported income. It’s hard to imagine any bureaucrat content to allow something like that to occur. And thus the power of the state would continue to ratchet upward.
Alisa,
I think that’s the difference between media – I was focussing on writing, you seem to be discussing conversation. I think though your point is a good one, that we can relate things to real world problems well enough. Problem is, I suspect many here, like me, prefer abstract thought…
True, Watchman – but writing and abstract thought that goes with it seem to get us only so far.
In any case, and to get back to the question JP asked in his post: UBI cannot be an improvement on the current welfare system, since by design it will be even more generalized (the clue is in the Universal bit), and even further removed from the individual needs of those it ostensibly purports to help.
Abstract thought is an excusee not to engage with real people, not a worthwhile end in this context, so it might not get anyone very far at all…
But I think that might be a very useful point about UBI – can we show how this will help any particular individual more than the current system, or a more targetted and intelligent system (so not government then). I suspect not, which would be the argument here.
To try to put this into practice, an ex-colleague of mine recently lost their job for drinking in the day. It appears (without her confirmation) she is an alcoholic (not a huge shock if true), and my information is she was doing her job really badly at the end, whereas previously she had generally produced OK work. So how would UBI help someone who appears to be on a downhill spiral like this – she would be given a regular amount to spend on fuelling an addiction, and would have no incentive to combat the problem: at least current welfare systems tend to require you to at least make a pretense of searching for work or getting medical treatment – the failures there are in the delivery not inherent. Someone who is making really bad decisions for some reason is hardly likely to be helped by UBI.
NickM, you missed the important qualifier in my previous post: you are not tech-dependent “in the sense of this discussion“. Sure you use computers, we all do; that’s minimal tech. What we’re talking about here, though, is technological advancement which eliminates whole categories of jobs. That’s not you (or your wife).
Lee, your “microeconomic” solution sounds great, in theory. In practice: no way; never work. Very few people will actually do that “hard and boring” work. Exemptions will become endemic. Single mothers with young kids will be exempt. So will the “elderly” (which will be progressively defined down). So will those with any sort of disability (which is already a serious problem, with people faking disabilities in order to get welfare benefits; imagine what it will become under a UBI scheme such as you’ve outlined). When Bill Clinton promised to “end welfare as we know it” it was partially changed into “workfare” (similar to your scheme). That only lasted a few years, because it was politically useful to reduce and eventually eliminate most such requirements. “Gaming” the system will be inevitable and unavoidable, regardless of how carefully it may be designed initially.
Watchman: Party on, dude!
Foreword: In my view, “BIG” or any of its aliases, siblings, cousins, etc., is immoral as hell and should be sent into the nearest Black Hole instanter.
Following speaks purely to the supposed meaning of the term “B.I.G.”
. . .
As I’ve seen the “B.I.G.” idea touted, the idea is that everyone (just who is “everyone” is a question) in the given country (say, the U.S.) will receive a check for $ X every month, year, or whatever. That’s a flat $ X, for everyone. WikiFootia:
https://en.wikipedia.org/wiki/Basic_income
[My boldface here, the Foot’s omitted.]
Matt Zwolinski’s lead piece in a debate or roundtable on B.I.G. at Cato Unbound includes this (again, my boldface):
https://www.cato-unbound.org/2014/08/04/matt-zwolinski/pragmatic-libertarian-case-basic-income-guarantee
Zwolinski also quotes from what he says is an amendment suggested in 2006 by Charles Murray:
—-As an aside, I can’t resist quoting this comment from responses to Zwolinski’s piece:
At https://www.libertarianism.org/columns/libertarian-case-basic-income , Zwolinski distinguishes among various types of, um, “help”:
SQOTD?
It’s very true.
“Society adapts” is so much less compelling than “last night, the poor buggy whip maker’s little girl died of starvation.”
Tarrou’s #4 pretty much sums up my practical* objection:
==========
4: The misreading of the hardness of the public. UBI requires that if a poor person with six kids blows their UBI on coke and whores, the voters look at the starving kids and reprise Scrooge: “Are there no poorhouses?” Maybe you think your electorate is that hard. I know mine isn’t.
==========
So what we will ultimately end up with is UBI and some variation on our current welfare system layered on top.
*The moral, etc objections have already been well covered in these comments.
James James writes: Citizen’s Income is an improvement on what we have. Complaining that one should not be entitled to a handout simply for existing is missing the point: one is already entitled to a handout simply for existing, but it’s a handout that gets withdrawn as you earn money, causing effective high marginal tax rates and decreasing the incentive to work. A Citizen’s Income/Negative Income Tax (they are precisely equivalent) avoids this problem.
“One is already entitled to a handout simply for existing”. So what? I am talking about what should be the case, not merely what happens to be the current default assumption (that welfare is a “right”, which it isn’t). UBI elevates the idea that receiving an income from the State (ie, one’s fellows) to even more of a “right”, and embeds even further the collectivist nonsense that underpins it.
It is not clear to me that a negative income tax/UBI or whatever avoids the marginal disincentive issue because, for example, if I receive a basic income regardless of whether I sit on my arse doing nothing or working, a lot of people who have few skills to appeal to employers/customers are going to do nothing. And there is the cost of paying for this, which I suspect will be greater than the current imperfect (as it surely is) arrangement.
How it is funded is a separate issue. You’re right that taxing robots is a tax on capital. But that’s irrelevant. It could be funded from existing taxes on income, capital gains, whathaveyou. It would be largely tax-revenue neutral anyway, because it would *replace* existing benefits. People who currently get benefits would get the same sum. People who currently don’t would get it, but would be taxed slightly higher towards the top end, so it would be revenue neutral. If a Citizen’s Income is funded from income tax, then it’s essentially reducing the high marginal tax rates on the poor, increasing their incentives to work, in exchange for a slight increase on marginal tax rates for the rich, slightly decreasing their incentive to work.
Agree; the point I wanted to make is that a lot discussions conflate the idea of taxing job-killing robots with UBI. Of course, UBI could be funded in a variety of ways. Income tax, by the way, is already pretty high in most Western democracies, and higher rates will hardly benefit the economic pie as a whole. There is also the fact that bringing such income in will also destroy incentives to save; and there is the issue of how the economics of this works when populations are getting older and the ratio of workers/retirees/those living on benefits is worsening.