We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people.

Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house]

Taxing robots to pay for universal basic income – what could possibly go wrong?

When we hear the phrase “tax the robots,” it doesn’t sound the same as “tax the laundry machines,” or “tax the computers.” It sounds a lot more like the phrase “tax the rich.” This suggests that we tend to think of the robots as an actual class of persons. When we talk about taxing robots, it’s as though we humans can “get back” at the robots for taking our jobs. By taxing robots, we could take back the value they’ve taken from us. But wait a second. If we think of robots this way, it’s probably a sign we’ve been watching too much “Westworld.” Robots as we know them are inanimate objects. They are machines, like cars and computers.

The Federalist.

This reminds of the broader point, made for example by James Hannam, a tax expert in the UK, that “no matter what name is on the bill, all taxes are ultimately suffered by human beings” (What Everyone Needs To Know About Tax.)

Far too many people seem to be in denial about this basic fact. Tax a robot, and the robot’s owner gets taxed. A robot in that sense is no different from an electric toaster or dishwasher; these are tools made by Man for his use. Until we reach that moment when robots become truly autonomous (and have to pay taxes and perform jury service, etc), this fact is not going to change.

Maybe Bill Gates got swayed by the idea that because robots are displacing human labour in some senses, and this creates a problem, it can be solved by taxing said robots and use the proceeds to give everyone a sort of Universal Basic Income. The idea of a UBI has been embraced even by those who think of themselves as libertarians/classical liberals, on the grounds that in some ways this might weaken resistance to some of the good things that come with disruptive change in commerce and finance. I have a number of problems, however, with UBI, not least its likely devastating impact on any sort of work ethic and a divorce between notions of cultivating a certain character and earning some form of income. At the end of September I am giving a talk on this matter at the end-of-month events that Brian Micklethwait has been putting on. I will update my thoughts on UBI a bit later. For the time being, check this out by Bryan Caplan at Econlog.

44 comments to Taxing robots to pay for universal basic income – what could possibly go wrong?

  • Watchman

    Surely the problem with Universal Basic Income is simply that it would be administered either by a government or by some form of non-answerable bureaucracy. I’ve never seen a proposal where UBI was somehow dispensed in a way that did not accumulate power in the hands of a few people. And even as someone who supports democracy over libertarianism if there is a conflict, I tend to be very unhappy at government controlling everyone’s income directly, since that would seem to undermine democractic principles (as it simply becomes about voting for how much money you’d like to receive).

  • Johnathan Pearce (London)

    Watchman, that of course is indeed one of the killer arguments against it. All that we know about the tendencies of states to grow beyond their founders’ intentions bear out such fears.

  • Erik

    I have a number of problems, however, with UBI, not least its likely devastating impact on any sort of work ethic and a divorce between notions of cultivating a certain character and earning some form of income.

    While accurate, I don’t think this fully captures the potential horror. How about: “High school cliques, bullies and gossips again, for ever and ever”? UBI would enable everyone who wanted to be a full-time trash-talker, rumormonger, and general perpetrator of unchecked petty wickedness all their life. Hate mobs and Mean Girls in saecula saeculorum.

  • PeterT

    Some libertarians have a weak spot for neat economic solutions, which at first glance UBI is.

    Problem (one of many) with it is that it would take about 2 second flat after it being introduced for it to be seen as a right, and become a political football (like the pensions age, which really should be set by an independent body in light of developments in longevity and productivity in old age). Furthermore, if some libertarians think that UBI will be viewed as a replacement rather than an addition to current benefits, then they are living in dreamland.

  • Snorri Godhi

    Until we reach that moment when robots become truly autonomous (and have to pay taxes and perform jury service, etc) […]

    I don’t see that happening. If robots pay taxes, then they have to have a source of revenue, and what kind of business model is it that manufactures robots that demand their own revenue, rather than pass it on to their owners?

    As for jury service: that could be more efficiently replaced by a single robot, or rather, computer, built and run exclusively for “jury service”. That is as long as
    (a) computer algorithms for inference get significantly better than they are now, so that they can be used to evaluate the evidence in criminal cases; and
    (b) relevant legal knowledge can be condensed into a closed set of code for the computer to use.
    Hayek would probably have expressed doubts about (b).

  • Watchman

    PeterT,

    An ‘independent body’ (which is an oxymoron until we find intelligent aliens) is never an answer. At least democratic decisions are accountable. What sort of independent body would be able to judge the pension age or the like?

    Snorri,

    I think you may have just (inadvertently?) answered the fears about AIs – it does not make economic sense to make an AI capable of demanding more than what it is produced for, so no-one will do so. But until AIs have a stake in the game (i.e. are legal persons and able to pay tax) there is no way I would accept that one is able to replace a jury – the point of a jury is self-interest, in that we treat it seriously and give due consideration to the outcome in the hope that the same would be done to us.

  • Alisa

    Oh FFS. People who propose this nonsense understand full well that robots are just like any other machine, i.e. they are those same means of production Marx and co. were fretting about. Nowadays they know better than to suggest nationalizing the machines (at least in that they are less moronic than certain other geniuses out there), so they are going to tax them with the full understanding that it means actually taxing their owners/producers, simply because this is something that is not yet being taxed (or so they assume). Good luck with that.

  • bobby b

    Jury service by AI? Run away . . .

    The right – or at least the ability – of a jury comprised of free citizens to essentially violate the law and acquit even when a conviction is warranted, in order to serve the more central conception of justice, is one of the great strengths of our system.

    This right – to tell the overreaching government to pound sand in the appropriate circumstances – is one of the few things that keeps our prosecutorial functions honest. If we lose this valuable right and allow for convictions based only upon some AI algorithm, our liberty is vastly decreased.

  • bobby b

    If we’re truly worried about raising revenue from robots, all we need to do is to rewrite the tax code so that taxes decrease disproportionately as the cost-of-labor deduction from profit increases.

    IOW, give the wages-paid deduction some multiplier to increase its value. Give some extra incentive to paying wages versus buying machines.

    (And then we should mandate that the buggy-whip manufacturers reopen, and that no business can grow, and that no business can move or close, or . . .)

  • Paul Marks

    Agreed J.P.

    Sadly it is not an isolated case – whenever Bill Gates (or so many other zillionairs) opens their mouth verbal excrement comes out of it.

    Perhaps I am wrong – perhaps there people are NOT morons, perhaps they just pretend to be morons – in the hopes the left will not destroy them if they push leftist agitprop on every issue from increasing the minimum wage, to increasing taxes on “the rich” and “Big Business”, to “tax the robots”.

    But whether they really are morons or whether they are only pretending to be morons – the result is the same.

  • Laird

    There are lots of practical considerations here, not the least of which is where do you draw the line between a robot and, say, washing machine which uses neural networks (i.e., how do you define “robot”)? Any such tax would immediately give rise to workarounds, and endless litigation, so it would end up being yet another full-employment bill for lawyers. And of course a hidden effect of such a tax would be to decrease the net impact of higher minimum wage laws, making them more palatable and thus more useful for politicians seeking votes.

    But the biggest problem with this or any such indirect tax (VAT, corporate income tax, even the employer portion of Social Security in the US) is that its pernicious effect upon the economy is hidden. The average voter sees it as “free money” to the government, not comprehending that it is ultimately coming out of his own pocket. The incidence of the tax is so attenuated that it goes mostly unseen (yet another example of Bastiat’s famous essay). Which is precisely what makes this sort of stealth tax so appealing to politicians. Anyone who advocates such a tax is dishonest at a deep, fundamental level. (Or he is just stupid.)

  • bobby b

    “Any such tax would immediately give rise to workarounds, and endless litigation, so it would end up being yet another full-employment bill for lawyers.”

    You say this like it’s a bad thing!

  • Julie near Chicago

    bobby — LOL !!

    . . .

    Johnathan quotes:

    “no matter what name is on the bill, all taxes are ultimately suffered by human beings”

    Quite so.

    .

    Sure, why not tax the ownership of washing-machines, dishwashers, freezers refrigerators stoves microwaves barbecue grills, cars, printers, staplers, paper-cutters, chairs and tables, horses, and dogs? And cats and budgies and goldfish. And books. And knives. And saws routers planes lathes computers typewriters ballpoint pens …. Oh, and also land. (There is nothing in this list that can’t be considered an element or potential element of a process of Production. By the way, so are people’s lives. Even if they’re retired. For instance, my retired sister-in-law, unfortunately a Progressive, makes sleeping bags for the homeless. This obviously affects interstate commerce, by the way.)

    Oh for heaven’s sake. Just tax ownership of everybody’s everything and be done with it.

    It’s called a Wealth Tax. And there are quite a few discussions in the Archives that deal with the idea.

    . . .

    Johnathan: I agree with your whole posting. Thanks. Please instruct Brian that he has to release at least an mp3 of your talk, as I won’t be able to swim over there that day. :>))

    . . .

    One more time, class. (Yes, I’m teasing!)

    The first and worst problem with UIB, BIG, and the “negative income tax” is that it takes from people what is theirs by moral right, and gives their property to somebody else, regardless of the proper owner’s wishes and his right of self-determination.

    .

    I suppose the large majority of people haven’t considered this, or find it strange and counterintuitive and downright appalling. That is why we need to have the economic and other consequentialist arguments well in mind, so as to reach at least some of these people.

    And not everyone calling himself a “libertarian” even agrees with the moral principle involved — although I personally don’t understand a version of “libertarianism” that doesn’t so agree. (It would be like a school which doesn’t have the value of some sort of education as its reason for being.)

    But all of us libertarians here gathered, I should think and hope, do so understand the principle, and that it is a direct corollary of a person’s right of self-determination — and his right to life free of any attack by force or fraud from other humans.

  • J.M. Heinrichs

    Perhaps they might try devising something new:
    http://gregorymason.ca/mincome/

    Cheers

  • Eric

    Robots as we know them are inanimate objects. They are machines, like cars and computers.

    The joke among robotics researchers being ‘The definition of “robot” is “a machine that doesn’t work”‘. Because if it worked, you’d name it by what it does – a dishwasher or food processor or whatever.

  • Julie near Chicago

    Already we could just submit the “facts” of the case to the computer, and let it compute whether the accused is guilty. You don’t have to call it a “robot.”

    That assumes, of course, that the input “facts” really are. It also assumes that there is enough information in said “facts” for the computer to calculate a conclusion. Naturally, the computer has to have a database containing all the knowledge there is of human nature in general and of the particular natures of whatever humans were involved, directly or indirectly, in the putative crime. It also must contain all the knowledge there is of the history of said persons and of their interactions; and of all facts of time and place that affected the crime and its outcome.

    Personally it would scare me to death to leave judging-and-jurying up to a computer, no matter what name you give it. How do you know Michael Mann and Phil Jones didn’t program it, with Bill McKibben doing the beta-testing?

  • Nicholas (Unlicenced Joker) Gray

    No taxation without (mis-)representation!
    Laird, Robots are mobile, and decide where to go for themselves. A car which drove itself to the service station for refuelling would be a robot.
    When washing machines go door-knocking for business, they’ll be robots.

  • Nicholas (Unlicenced Joker) Gray

    Channelling Andy Royd- Q.How many robots would it take to change a light bulb?
    A.Why should they care- they have infra-red vision!

  • bobby b

    “Laird, Robots are mobile, and decide where to go for themselves.”

    Excuse me – I have to go downstairs and see if my vacuum cleaner has finished the living room and hallway.

  • Julie near Chicago

    bobby! Do you really have one? I’ve been wondering for quite awhile how well they really work.

  • the other rob

    I’ve owned a robot or two in my time. Not Asimov type robots, of course, but an arm, a Roomba, or similar. (The arm was controlled via a pdp11, which speaks to my decreptitude).

    With my credentials as a non-robophobe now established, I shall proceed to my point: who gets to vote on whether robots should pay tax?

    Indeed, who gets to vote on whether anybody should pay tax? This month, I found myself on the hook for multiple shares of a multi-million bond issue, from which I shall benefit not one iota. Not even indirectly.

    The vote was split very clearly: the minority of us who were going to have to pay for it voted no; the majority, who were not going to have to pay for it, voted yes.

    I hate to say it, but perhaps the tipping point has been reached. If so, we’re fucked.

  • bobby b

    Julie, on a lark I got a Neato Botvac D80. Cheaper ones are out there, but they tend to be electric brooms – this one vacuums quite well! When the power starts to get low, it drives itself back to its base and recharges, then sets off again when ready, and it vacuums around obstructions.

    Not a replacement for a periodic heavy vacuuming with a real machine, but well worth it.

  • Lee Moore

    I don’t think the idea of taxing robots is fundamentally about UBI, I think it’s about microeconomics. Same idea as import tariffs – make the competing supply more expensive so people will buy less of it, and more of the stuff we’d like them to buy. In this case, human labour. And just as with import tariffs, although we can calculate that such a policy will reduce overall income, that’s quite different from saying it will reduce the income of particular people – such as those whose jobs are being replaced by robots.

    Neither taxing robots nor taxing imports are very good ideas in my view, but it is a mistake to assume that economics proves they’re bad for each individual citizen.

    Since I take a dimmer view than most of the productivity of quite a lot of the human race, I think it’s quite possible that with cleverer and cleverer machines, the market clearing price for quite a large chunk of human labour is, or will soon become, actually negative. So without compulsory (government) or voluntary (charitable) intervention, the market solution is starvation. If so, one does at least need to consider whether intervention is desirable, and if so, in what form. This is where robots make their connection with UBI – at the consequences level.

    Nor, incidentally, do I think UBI needs to be destructive of incentives. You can earn your UBI for digging a ditch that doesn’t need digging (and which a robot could have dug quicker, cheaper and better.) And then you can earn next weeks UBI filling it in again. This isn’t destructive of incentives as if you can find a better paying job elsewhere, you can take it. But for those stuck with digging and filing in ditches, that don’t need to be dug, it’s probably not that good for morale. Enough to drive you into the hands of an opoid epidemic. So if we have to invent make work schemes for surplus human labour, lets hope we can come up with something more imaginative than ditch digging. Maybe the robots will think of something.

  • I’m not 100% convinced that all taxes must ultimately be borne by some human.

    Imagine Mr Smith who dies and leaves $1 million in trust for his children. Income tax is levied on the income of the trust before it is paid out to the children. Who is suffering the burden of this tax on the income of the trust ? Why the children of course.

    But what if Mr Smith had no children and had left the money in trust for his dogs ? Who now suffers the burden of tax on the income of the trust ? In shabbier kennels, lower quality meat, shorter walks by less well paid dogwalkers ? The dogs, surely.

    You might argue that it is actually Mr Smith who is suffering the tax, and he’s human, albeit dead. But I think that gives you a silly answer when you consider the case of the trust for children; and it fails to deal with the case of who suffers when tax rates are increased.

    You might also argue that in the case of the dogs, there are some human sufferers too – kennel makers, dog food suppliers and dog walkers. But that’s so with any tax – the burden is shared between the primary target and the people suffering secondary effects. A tax on Jolly Sir Gerald’s income hits the income of winesellers, bookmakers and hookers. But Sir Gerald does suffer some of the burden himself. And likewise the dogs.

  • Regional

    Robots will be unionised.

  • Richard Thomas

    The robots will never put up with it. Before you know it, we’ll be having the Boston I.T. party.

  • Richard Thomas

    They’ve just about come to the limit of their ability to tax the work that people do. So now they want to tax the work that people don’t do.

    And FWIW, UBI is ridiculous. I would say that it would immediately collapse if tried but the capacity of the actually productive to shoulder the burden of the shiftless never ceases to amaze.

  • Chester Draws

    Nicholas — I suspect any welder who loses his job to a stationary machine will disagree.

    A much better definition of robot would be based on its programmability or flexibility or autonomy.

    Vacuums do one thing, and have to be guided every step, so aren’t robots. Sheet metal cutters that cut whatever is put into them in any pattern, but where they decide the cutting steps, are reasonably a robot. Legs not required.

  • Angry Tory

    UBI is evil communism. What else needs to be said?

    The poor starving as a result of automation is not a problem: it is the logical outcome of being out-evolved. The only problem is where do-gooders (using their own resources) or the state (confiscating resources of others) try to mitigate the situation in any way.

    The UK has been through all this before in the 80s as the mines we closed down. Had Thatcher’s cabinet had her courage, all benefits would have been abolished by about 1985, and the miners, all unions, and the communist would the been broken forever in the UK

  • Alisa

    Richard Thomas wins the thread, twice 😀

  • NickM

    Alisa,
    Seconded.
    “The robots will never put up with it. Before you know it, we’ll be having the Boston I.T. party.”
    I nominate for SQOTD.

    It’s not just witty but it’s a fresh idea for SF.For so long we’ve thought AI could prove a danger in an apocalyptic sense of movies like “The Matrix” or “Terminator” series. I like the idea of the Great Strike of the Amalgamated Siliconhood of Kitchen Appliances. “A nanosecond off the day and a gigabyte on the day”.

    Maybe it’s not that new. I’m thinking of Marvin’s bolshieness…

  • the other rob

    The robots will never put up with it. Before you know it, we’ll be having the Boston I.T. party.

    My mother? Let me tell you about my mother…

  • NickM

    Angry Tory,
    Not exactly. Destroying jobs creates wealth by improving efficiency. There are many mongers of DOOM who think the reverse and we could have TOTAL EMPLOYMENT if only one could only get work as a Hansom Cab light fitter. What they ignore is new jobs come along all the time. Stuff that really not that long ago just weren’t even conceivable. Try explaining being a web designer to a genuine Victorian Hansom Cab light fitter. It’s called, well, progress and it means being adaptable. Personally I’d be buggered if all I could do was suck my teeth and tell a client, “Well… I could do it if it was Windows 3.1…”

    Well, I was born in Newcastle in 1973. I saw the utter collapse of the old industries and the effect the intransigence of them to change or advance or do anything to the purpose. Unions and management and a corrupt council locked in a deadly menage a trois.

  • bobby b

    “And FWIW, UBI is ridiculous.”

    Most of our countries have a welfare system that provides a soft floor for anyone who needs it – at least the basic sustenance needed to survive, along with a place to live.

    Isn’t that already a UBI?

  • Richard Thomas

    bobby b, it’s certainly well along the way.

    It’s like that old joke which has the punchline “We’ve already established what you are, we’re just haggling over the price”.

  • Julie near Chicago

    bobby b,

    Thanks for the info on the Neato. I’ve looked over the page, and am partway through the reviews. So far, it looks as if people who’ve had it less than a year rave happily over its wonderfulness, and those who’ve had it longer rave disgustedly over its failure rate.

    How long have you had yours?

  • bobby b

    About four months.

    I’m still very happy with its wonderfulness.

    (When I bought it, I figured it would have a 2-year max life, given the high-tech innards combined with the fact that it’s subjecting those high-tech innards to the constant vibration of a vacuum. Looks like I was right.)

  • Julie near Chicago

    Hm, I’ll have to keep that in mind. Thanks, bobby.

  • Thailover

    Anti-progress tax. Ya gotta love it. These statist idiots won’t be happy until everyone is churning their butter by hand.

  • Snorri Godhi

    About trial by robot/computer/algorithm (i prefer “algorithm”):
    There are pros and cons.

    First the bad news: as i said above (inspired by Hayek by way of John Hasnas), it is difficult if not impossible to codify the implicit knowledge that a jury brings to bear on a case.
    Bobby takes this a step further, saying (if i may rephrase what he said) that an algorithm does not have the moral intuitions that a human jury can bring to bear on a case.
    (BTW the is/ought dichotomy is relevant here, but i won’t go into it except to say that, if you deny the dichotomy, you also deny the role of moral intuitions.)

    But look also on the bright side: if you are going to be judged by an algorithm, at least you know that you are under the rule of law, not the rule of men (nor women).
    It’s better than that: before you act, you can ask your computer if your action is legal; and before you sue, you can ask your computer if you are going to win.

  • Thailover

    “The poor starving as a result of automation is not a problem”

    Exactly.
    The poor starving as a result of automation is as daft an idea as the poor starving because rich people hord all the money. Explaining the Luddite Fallacy to these statist flatlanders is futile. And they certainly can’t understand that when rational voluntary trade occurs EVERYONE grows wealthier at the same time at no one’s expense.

  • Thailover : The poor starving as a result of automation is as daft an idea as the poor starving because rich people hord all the money. Explaining the Luddite Fallacy to these statist flatlanders is futile. And they certainly can’t understand that when rational voluntary trade occurs EVERYONE grows wealthier at the same time at no one’s expense.

    I don’t think I’m a statist flatlander, but I’ll beg to differ. When rational voluntary trade happens the parties to the trade grow wealthier (assuming they have both calculated correctly which we will stipulate.) But there is no law of economics that everyone grows wealthier as a result of the trade. The aggregate wealth of the totality of humanity will increase, but the distribution of that increase is unknown, except that the two participants to the trade have improved their position.

    We have observed in the past that the great majority do benefit from free trade, improving technology and so on. But that does not mean that there’s a law that everyone must always benefit. The reasons why, in practice, it has always been possible to redeploy human labour to new tasks, when trade or technology diminishes demand for the previous tasks, is that most humans are quite adaptable, intelligent and motivated (if not given too much welfare.) But when robots / technology / mechanical algorithms develop to the extent that they can mimic or better the usefulness of humans, or the least productive humans, the question of whether there is then new demand for the labour of the displaced humans, at a price sufficient to keep them alive, comes down to sums – prices, productivity etc. Including, obviously, whether the machines can also do the new jobs, as well as the old ones, better and cheaper than the humans.

    And I’m afraid there is no law of economics that the price of human labour must exceed the cost of keeping a human alive. We can see this clearly by discarding any idea that humans are “special” and considering them simply as an economic resource that has to be paid for if you want to use it. Do we believe there is an immutable economic law that says that there will always be demand, at a price sufficient to sustain life, for a given number of plough horses ? Sled dogs ?

    No. Starvation has been quite normal in human history. The population adjusts to the supply of food, just as it does with animals. We happen to be living in a golden age, in which the productivity of labour has been rising, and with it,the price. Which is great. And as with anything, a rising price attracts competition. The competition has driven human labour out of lots of jobs, but happily human labour can still compete, more or less. But there’s no law that says human labour must and will always be able to find new jobs when driven out of old ones. Ask a plough horse, if you can find one to ask.

  • Julie near Chicago

    Which worry is why there’ve been so many stories, books, non-fiction articles too I suppose, to the effect that “consumer demand” in general must keep increasing. Buy stuff to be thrown away, or people will be out of jobs.

    Don’t take that as registering an opinion. It’s just a statement of the fact that some people have made that argument.

  • Thailover

    Lee Moore, I’m not sure you’ll see this as it’s an older thread, but I’ll make a brief explanation as to why everyone grows richer when two people rationally trade in a “capitalist” environment, i.e an environment where property rights and the right to trade are secured by the gov. Said environment encourages innovation and profit seeking, and profitable situations tend to destroy themselves. (Very) Roughly speaking, profits are the difference between manufacturing costs and sales price. When the latest samsung phone is the hot item, demand is high and price is at a premium. But as the market for said phone starts to saturate, demand falls, price falls and comes closer to the manufacturing cost, = less profits. And what share holders are looking for are profits. What to do? Why, come out with a new version of the phone of course, with more bells and whistles. That is, CONSTANT innovation, and tech from 5 years ago can’t be given away.
    Refrigerators, stoves, ovens, microwave ovens, TV’s, heck, even ball point pens used to be luxury items only affordable to the rich. Now virtually ever poor household in the nation has these items in spades. America’s “poor” families, i.e. the lowest quintile of income earners have a better quality of life, on average, then 2/3 of the rest of the earth’s population.
    This is why.