2009 August 09 Sunday
Gregory Clark Expects Unemployable Lower Class
Gregory Clark, historian and author of the noted book A Farewell to Alms: A Brief Economic History of the World, has an op/ed in the Washington Post on his expectation that smart machines will make the lower skilled human workers unemployable.
The battle will be over how to get the economy's winners to pay for an increasingly costly poor. Last weekend Treasury Secretary Timothy Geithner and Lawrence Summers, the director of the White House's National Economic Council, refused to rule out raising taxes. Despite the White House's subsequent denials, this may be an early acknowledgment of an inexorable trend. In a future with higher taxes, the divide between rich and poor would be the central economic challenge.
Lower IQ people will become irrelevant as they are increasingly replaced by robots and other machines.
Even outsourcing of phone calls to very cheap Indian call-center workers is getting replaced by computers.
I recently carried out a complicated phone transaction with United Airlines but never once spoke to a human; my mechanical interlocutor seemed no less capable than the Indian call-center operatives it replaced. Outsourcing to India and China may be only a brief historical interlude before the great outsourcing yet to come -- to machines. And as machines expand their domain, basic wages could easily fall so low that families cannot support themselves without public assistance.
If Indians getting paid much less than American minimum wage can't compete what to do about our growing lower class?
Clark expects unemployment on a massive scale.
With the march of technology, the size of a future American underclass dependent on public support for part of its livelihood is hard to predict: 10 million, 20 million, 100 million? We could imagine cities where entire neighborhoods are populated by people on state support.
More than half the American population has an IQ of less than 100 and that percentage is growing. We really need to stop letting people enter the country to do manual labor. The people stuck in the lower classes will feel their decline in status. Whether they are given make-work jobs or welfare they will feel frustrated and angry at their low status and lower power.
Also see my post Robots To Cause Mass Unemployment Of Low IQ Workers.
Update: In the comments someone who doubts the eventual unemployability of most low IQ people points to sweeping as a job still requires human labor. I have a couple of answers for this that you can buy right now:
Imagine what cleaning robots will do 20 years from now.
Well if anything close to the singularity hits in the near future, everyone's fucked in terms of employment and relevance, not just the idiots.
Robotics + AI is still extremely poor for anything other than do the same simple thing over and over again. I don't see any great leaps on the subject anytime soon.
I think computers are going to replace mangers first. Computers can already schedule and track things better than people. Most of the manger training for most places is about getting the person to run the store by the book and keep the screw ups to a minimum. You can take a average IQ person give them a head set telling them what to do and replace the manger with that setup. You would get more consistent performance and you could pay them a who hell of a lot less.
Classical economic theory dictates that the amount of 'wealth' in any given society is a function of the 'productivity' of that society.
Supposedly labor is the 'irreducible cost' and by reducing the labor content of goods and services as close to zero as possible, the cost of the product tends to zero, thus, the product is more affordable and consumers 'wealthier'.
This is the whole story of economics from the agricultural revolution onwards from smelting iron with coke, from the sewing machine, the power-loom, the steamship, the railroad, the automobile, artificial fertilizers etc etc onwards - the trend has always been to reduce the labor content of goods and service and thusly we are materially 'better-off' than our ancestors.
Now, supposedly jobs will decrease as labor is cut-out - but (supposedly) this is an economic fallacy.The utilisation of labor is a function of the 'wealth' of that society.Individuals who now cannot afford certain goods and services will now be able to afford them, opening up new avenues for labor, how many metal machinists were employed 200 years ago?, how many electricians, air-con engineers, printers etc?
If you think that economic theory is actually worth a damn (and I don't), then if the theory actually worked an the USA's GDP actually grew at the optimum level (supposedly the name of the game) then I gurantee you not a single dummy would be out of work (with a well paying satisfying job at that), unless he chooses to remain so.
And I'm not even talking of Chinese rates of 10% GDP growth, but more like 5% per annum.Supposedly rationalization driven by technology drives this growth.
People with money always want to live in better homes - the construction industry - most trades cannot be further mechaanized soaks up huge amounts of grunt labor.Likewise consider the wealthy metropolis of the future, just think of how a well planned city can be beautified into an earthly paradise by employing teams of professional gardeners?
Think of Switzerland - a very wealthy technological society that imported grunt labor by the milion.Think of Dubai - who built all of those skyscrapers - my point is that labor is the major irreducible factor of wealth and what causes unemployment of dummies i conomic stagnation NOT substitution by hi-tech, which in fact increases wealth.
The fact that SoCal imoprted alot of irredeemable trash to do its grunt work in happier times is another story.
The people on the bottom may feel angry and resentful, but they will not be powerless - they can still vote for politicians who promise them lots of goodies, such as free health care, free food, free housing, etc. This is what the Democrats are doing right now. To be honest, the Republicans under Bush II tried it with boondoggles such as Medicare Part D and No Child Left Behind, but now they seem like pikers in comparison to the porkfest and vote-buying of Obama and his friends in Congress, who are trying to get their own "entitlement" to a bunch of private jets so they don't have to wait in airport lines as their constituents do. This is the "bread and circuses" model that the Romans used to pacify their population - how'd that one work out for them?
"This is the "bread and circuses" model that the Romans used to pacify their population - how'd that one work out for them?"
Well, there is no other option besides bread and circuses. BTW, Sweden has the bread and circuses model, and it worked.
Maybe it would be easier to control the masses by giving them make-work jobs. Why do Austrians hate make-work jobs? They do not create wealth, but they transfer it, and suppress outrage since it gives people the illusion of doing something productive. I like make-work jobs, it is my idea of bread (the wages from the job) and circuses (the job itself). I hope the US tries to quell the masses with bread and circuses instead of outright oppression. (Too bad, I do not like the current form of "circuses" ... I never watched a full episode of "American Idol")
I do not see any meaningful way to increase employment besides make-work jobs and reducing competition via import certificates and restricting immigration. Make-work jobs are a transfer of wealth because it transfer income from those who pay that taxes to those who get the make-work jobs. I hope the US chooses the bread and circuses path instead of oppression. (Of course, as I implied before, perhaps the circuses are already too effective at suppressing political interest. People are more concerned about American Idol instead of politics.)
Call me skeptical.
People have been predicting since the 1960's that computers and robots were going to replace all the workers in the "near future". 40 years later, it still hasn't happened. The AI keep saying every decade that "real" AI is "just around the corner". So far, they've been wrong as well. I subscribe to the "bio-singularity" concept, but I have never bought into the computer/AI singularity scenario at all. I don't see the smart machines that Gregory Clark talks about. I see computers and software that continue to be glitchy.
It is true that robots are becoming more prevalent in manufacturing, especially semiconductors and flat-panel display manufacturing. Indeed, I have programmed many of these robots myself. But I don't see the wholesale replacement of human workers with robots in vast sectors of the economy anytime soon.
But many corporations and special interest groups completely approved the policy of increasing the percentage of low IQ citizens because this created a lot of opportunities to take care of these groups by providing goods and services tailored to welfare and special entitlement programs. Until the de-industrialization of the United States has reached the current deleterious level, all of these machinations were tolerable, but now the situation became impossible.
Once again, the solution will start with the stopping of the trade deficit. Once the trade deficit is stopped, then the annual government deficit spending becomes far more possible, since the Fed would then have more freedom for replacing the government debt with cash by simply buying a lot more treasuries, and this allows the government to continue the same deficit spending the following year. This always works if the government deficit spending is in the order of 3 % of the GDP, which is the natural rate of growth, but it will not work if the foreign trade deficit becomes too large.
Apply commodity theory to humans. Cows, humans, cars or chopsticks - when the supply goes up and the demand goes down... And so does the price or value.
The "robots will displace the workers" is an example of bullshit argument for a very simple reason - it totally misses the point of Ricardo's comparative advantage argument about division of labor.
It is _comparative_, not _absolute_ advantage which matters. Even if robots (aliens, whatever) were million times, on average, more productive than mere humans, humans will still be better off having robots/alients/whatever around, simply because they are different - and so it will make sense for them to trade with humans (and humans will be better off in absolute terms because of the trade).
Every time somebody drag out this argument he simply shows his inorance in basics of economics.
I don´t know much about Gregory Clark, but if he was capable of formulating this argument, there is a solid case to be made that he is an economic illiterate
Since many others have debunked his idea in previous posts, I have only one point to make. If super-productive robots displace low-IQ workers, the wage of low IQ workers will go down, but as the total productivity of the economy grows, their REAL wage ( the wage in term of raw goods) go up. The stupid mexican was paid USD 100 and bought 20 pens with that money. Once a robot displaces the mexican, he´ll be able to find work only paying USD 50, for example
since the robot produces 10x more pens than the mexican, his real wage, in terms of pens, is now 100.
Note to Mr. Clark: We already have an unemployable lower class.
Make-work seems stupid. Mainly because someone in government has to decide what work to do. There is a market-based analog: a negative income tax rate on lower income brackets. I'm no economist, but this seems much superior and seems to accomplish the same thing in essence.
I don't know that comparative advantage necessarily puts this robots thing to rest. Y'all are being a bit abstract. What if in the future most jobs are related to repair and programming of bots, and it takes an IQ of 120 to do any good at all. Would there be a comparative advantage, today, of trading with a nation where every person has an IQ of 70? That is, 70 is not the population mean, but is the IQ of every individual. Probably only if the cost of carrying out trade is really low.
Does comparative advantage set a lower bound on the *price* of low-quality labor? What if employing some low-IQ person is indeed win-win - but only at $3 an hour?
Sure, $3 an hour may turn out to be a lot - enough to live on - in a robot economy with enormous supplies of goods. And yet, it might nevertheless wind up representing a far larger gap between rich and poor than we have today - and people might find that unacceptable. This is, after all, why latin American immigrant labor is (probably) a net economic loss: because the rich are going to wind up paying for their health care, use of roads, and other entitlements. If they were willing to come here and work for $3 an hour and take sole care of themselves, then at least on an immediate and purely economic level there would be a comparative advantage in having them. In practice, that is far from what happens.
On the other hand, I too am skeptical that there will be any real explosion of robots replacing low-IQ jobs. Certainly, some jobs will continue to be gradually replaced.
> Note to Mr. Clark: We already have an unemployable lower class.
Yeah, exactly - but I see that as a major support for roughly what he's saying. That class can only grow, whether quickly or slowly. Why would it shrink, unless $3 an hour in today's dollars becomes a living wage in 2025 and we allow firms to pay that wage despite the median salary having reached, say $150,000?
In reality, there will probably be life-saving medical techniques in 2025 that cost $500,000. The median person will be able to afford this. Bleeding hearts will of course demand that these be available to everyone. Not gonna happen for you at $3 an hour, unless the gov pays.
If you insist on being wrong so often you should try being less arrogant about it.
Sometimes one can have no comparative advantage. Transaction costs, transportation costs, or other costs can make anything one has to offer worthless or even of negative value. Given a large enough difference in general abilities some people won't have anything to offer. You can already see this today. There are peoples who make nothing that gets traded outside of their circle.
Some people are retarded, brain damaged from alcohol or drug abuse, schizophrenic, or otherwise disabled in ways that make them unable to produce anything of economic value. The threshold of needed competence for creation of goods and services of economic value will rise as more and more of the goods that the least able can produce will be manufacturable elsewhere for less.
Since some people are already incapable of producing anything of value In the future as the value of human manual labor falls a much larger fraction of the population will drop below the threshold of being able to make goods and services that have value.
Artificial intelligence is not needed to automate many service jobs. The Japanese ramen cooker is not artificially intelligent. Automated cow milkers are not artificially intelligent. We do not need artificial intelligence to automate the process of making a hamburger. We do not need artificial intelligence to make Roomba vacuum robots.
I started writing a reply to averros, but it got too long so I turned it into a blog post:
The very short version - the opportunity cost of taking a robot off of a job can never be greater than the cost of the robot, because you can always manufacture/buy another one (whereas you can't buy, say, another doctor - the whole process of making one involves expensively and lengthily raising a kid and training it). When the cost of a robot is less than a subsistence wage, a human can't compete with the robot in any field where the robot is competent. So comparative advantage doesn't really matter here.
I agree sentient AI is not necessary to replace a lot of workers. However, that is not the issue.
The big problem with automating service jobs is in vision recognition and dexterity. The biggest problem in AI right now is showing an image to a robot getting it to recognize a particular item from a background collage. This is called visual recognition and it is still one hell of a problem in robotics. This problem has not be solved as far as I know.
Dexterity and movement is not quite so difficult, but still difficult enough. Dexterity is being able to pick something of arbitrary size up without crushing it and putting it somewhere else without dropping or breaking it. Wafer and other substrate handling robots go about this in a very fixed pattern way. I think movement may have been resolved. Anyone who has seen the Boston robotics "dog" can see that mobility is no longer a problem. I think "spider" robots are the way to go.
Nonetheless, improvements in this technology will continue on an incremental basis. Instead of sudden massive unemployment in the next 10-15 years, there will be a gradual erosion of low-skill work over the next, say, 30-40 years. The rate of change will be gradual enough as to allow society to adapt to it.
--Instead of sudden massive unemployment in the next 10-15 years, there will be a gradual erosion of low-skill work over the next, say, 30-40 years. The rate of change will be gradual enough as to allow society to adapt to it.--
Whatever keeps you whistling past the graveyard....
Have you ever looked at unemployment rates by level of education? How about average wages for manual laborers since 1970? What Clark is talking about is already happening.
Think about all the specialized dog breeds that were bred for various commercial purposes. How many dog breeds still get used for their purpose? Dogs supposedly have the IQs of human babies aged 2-2.5 years. For many purposes their value has hit 0. They only have appeal as pets.
One way to deal with the vision recognition problem is to avoid it. Design what you are trying to do in a way that avoids the need. That isn't always possible. But it is possible sometimes. Or design covering or backgrounds that make recognition easier.
Look at the Roomba. It doesn't need to visually recognize a couch or chair.
Gradual progress: This does not prevent some people from becoming worthless in the labor market.
Many factors contribute to increase the Unemployment Rate. The average wage for manual workers is stalled because America is being flooded with manual laborers
let´s go back to the doomsday commenter who imagined a robot whose "wage" was below the subsistence wage. People tend to believe that in this case the whole Ghetto will be unemployed. What that commenter forgot to notice is that in such case the subsistence level will decrease in nominal terms but remain cosntant in real terms
You pay the minimum wage (USD 100) for Jose to clean your pool. With that wage Jose avoids starvation. But Satoshi, the new robot, can do the same service and Satoshi´s cost is USD 90´per month. But if Satoshis everywhere make the economy 10% more productive, if Jose accepts a job in your neighbour with a wage of USD 90, he´ll get the job and maintain his living standard. Because the economy is more productive, with USD 90 you are now able to buy the same basket that costed USD 100 before Satoshi
If the economy grows and Jose´s wage increases to 95 USD but Randall PArkers allows Jesus to immigrate, then Jesus will lower Jose´s wage. Jesus will also reduce the incentives for Americans to create robots.
for all that matters, the population of dogs is even more dependent on the Welfare state than blacks, both based on taxing whites. The difference is that dogs get it through the will of the white majority
and dogs are not a factor of production. Human Labor, Land, Natural Resources and Physical Capital are. Dogs do not generate wealth. What has been predicted for the last 200 years is the labour singularity when marginal workers will generate less wealth through their labour then the wage necessary to avoid starvation
people tend to think of productivity is increasing (which is correct) but the threshold to avoid starvation is fixed on stone (which is wrong). once you realize that as productivity increases the threshold in nominal dollars to avoid starvation will be lowered, Clark´s analysis stops making sense
One way to deal with the vision recognition problem is to avoid it. Design what you are trying to do in a way that avoids the need. That isn't always possible. But it is possible sometimes. Or design covering or backgrounds that make recognition easier.
This is true, but an unsolved vision recognition problem will act as an inhibitor of adaption of robotics in the marketplace. Much of the work that employs unskilled or semi-skill workers is that which requires vision recognition.
Ah, the comparative advantage deniers:)
Well, if you have mathematical inclinations you may have noticed that the law of comparative advantage does not require a lot of assumptions (despite the numerous attempts of pseudo-economists to claim that it does). What it says, basically, is that if two economic actors are different (have different shapes of productivity profiles) they can both benefit from trade.
It does not require absolute advantage, it does not care about reproductive rates of actors, it does not care about transaction costs. What the transaction costs do is simply *prevent* some potentially advantageous trades, which still leaves the actors no worse than they were before.
The "robots vs humans" scenario could be structured around another economically illiterate hobgoblin - the scarcity of resources and competition for them (it is illiterage because *all* material resources are fundamentally scarce, the fact often missed by Malthusians and their followers - and because price signals caused by increasing scarcity do cause more efficient use of scarcer resources). As long as there's no aggression involved, the robots are still no danger for humanity - if they start depleting resources, the humans (who initally own all resources on Earth) will keep raising prices, forcing robots to provide more and more of their products in exchange for some raw materials. Or the robits will go out and start expoiting vastly bigger resource pool in the asteroid belts and clouds - incidentally, making a portion of these resources available to humans.
There are only two plausible scenarios in which robots will take over completely - one, if humans will be stupid enough to start a war against robots, replacing economic cooperation with massive aggression, and another - if all humans will decide to *become* robots (brain uploads, progressive cyborgization, etc) for the benefits (like de-facto immortality) it confers.
How then would you account for there being any chronically unemployed at all, today?
Contra your post, there are certain transaction costs that pretty much apply to *all* employment. One transaction cost is the risk that a hire might steal from the business. Why do you think it's hard for convicts to get a job.
Other risks include lawsuits and property-destroying accidents.
Also, are you assuming there's no minimum wage?
In the auto industry the more automated factories produce higher quality goods than the less automated factories. Robots increase precision and quality. They do not just cut costs.
In the semiconductor industry and in some other industries robots do things that humans are incapable of doing. Some types of products could not be made without robots. Humans do not compete for these jobs. Humans are unable to compete for these jobs.
If the percentage of goods best made by robots continues to grow then there's no lower human wage that'll get humans back in doing those jobs.
I understand that if Bill and Susie stop using Jose and Luis to do their lawn and use a future robot to do it for less money instead then Bill and Susie have more money to spend on other goods. But we can't assume that those other goods will get made by Jose and Luis. If Bill and Susie desire to spend their money on more goods containing semiconductors then Jose and Luis might just be out of luck.
That's the problem. You can't count on freed up money to flow some other way to the people whose jobs got automated out of existence.
Now, you can claim that the cost of food for Jose and Luis will go down due to automation. So maybe their wages won't go down any faster than prices. But I do not think we can count on it. If demand for their labor totally evaporates then only free goods will help them.
Look at rising inequality. This amounts to upper class people buying more stuff from each other and less from lower class people. The lower class people become as irrelevant as horses that no longer get used to carry food to market. Again, the demand for other species has evaporated in many instances. The IQ level where demand for labor stops will probably keep rising.
But dogs are a factor of production. They are laborers for human enterprises. They are still heavily used for sheep herding and other livestock herding. They are also used for security work. They guard construction sites. They apprehend criminals for the police. They are much less a factor of labor than they used to be. They aren't used as much for hunting for example. They aren't used as much for rodent control. Medical researchers are trying to duplicate the functionality of a dog's nose to find cancer rather than just use dogs. I'm sure that'll eventually result in less use of dogs for bomb sniffing as artificial sensor devices displace them.
Work animals show us the future of lower IQ humans.
Here are just some transaction costs for using labor:
- finding possibly appropriate labor. I've spent a lot of hours reading resumes, interviewing people, discussing interview results. Companies spend to fly people to interviews, pay moving expenses, etc.
- instructions and supervising.
- theft, fraud, other illegal activity.
- law suits.
Robots won't sue, steal, or cause many of the other problems above.
But Randall, as I said before the real cause of unemployment is economic stagnation and slow growth.
In any advanced society productivity increases year-o-year resulting in a certain displacement of workers.In order to soak up the excess the economy needs to grow at a certain rate each year, the fact that economic theory is largely bullshit - China's massive growth hasn't dragged up America's growth is another story.
Consider this fact.Prior to the invention of the sewing machine in the mid 19th cetury, millions uon millions of women in virtualy every nation on eath were employed as seamstresses.In fact apart from prostitution and domestic service it was the only 'job' that women held.Women would take home cut pieces from a tailor and sew them together manually at home for a few cents a piece.
The invention of the sewing machine and thus textile factories destroyed this trade overnight, and the price of clothing dropped substantially (previously most people only owned one of two shirts, trousers or coats).The increse in productivity that occurred 150 years ago vastly outmatches anything we see today with robots or computers - BUT to labor my point, the excess labor released was soaked up by new industries that took advantage of the general increase in disposable income.
This process has been ongoing for hundreds of years, there's nothing new about it.
Look at Dubai - lots of wealth - lots of dummies employed.
If SoCal had Dubai levels of wealth, think how clean the streets would be.
- The fact that SoCal imported a load of trash has possibly poisoned your mind.
Dubai is a particularly poor example, because it produces essentially nothing that it uses. The vast majority of Dubai's materials, machines, electronics, and even food are imported. All the grubby details of those industries (outside of finance, construction, oil trading and tourism) are not less of a problem just because they are elsewhere.
If robots suddenly replace the entire workforce because they are twice as productive, the entire workforce could be re-hired by half their former nominal wage and everyone would be better because output would have grown 50%
Now you worry about a sectoral impact. Your last example, of the auto-industry, is outside the scope of Clarke´s article and could be considered off-topic
Since you cited a highly technical field, imagine 10% of workers producing 30% of the GDP. Now all those workers are fired because robots can produce 5x as much for the same wage. This economy has grown 120%. The output was 100, but now the robots are producing 150 instead of 30, so the GDP is now 70 (the output of the low-productivity 90% workers) + 150 (robot output)= 220 instead of 100.
We have 10% workers unemployed now, but percapita income has grown 120%. What happens now? The 10% displaced workers will search for jobs in the other sectors of the economy. Since the economy is now 120% bigger, they could be re-hired by their former wages and those wages, in the more productive economy, would buy 120% more goods. But nothing assures that. What matters is that they could keep the standard of living accepting a reduced wage while the other 90% workers would see unambiguous benefits
You are explaining theory as if we do not understand theory. Look, I've read Ricardo. I get it. I also see circumstances under which demand for types of labor can plunge to 0.
Animals provide an example of what we are talking about. They aren't as smart and many types of animals that used to be do productive work there is no more productive work for them to do.
Try considering the possibility. Imagine a world in which smarter people would rather use robots for anything that 80 IQ people do now. You think that is impossible?
Your 'horse' analogy doesn't really stand up.Presumably before the horse was domesticated (surprisingly late in the day), then to carry out any meaningful economic activity (building structures worthy of the name, communicating and trading with others, shifting any type of heavy material, building ships - yes, ships need very heavy logs to be dragged for miles etc etc), needed teams of 'dummies yoked together laboriously dragging the said factor of production many miles.
Of course, once the horse substituted the brute labor of the dummies, the dummies were freed up to do something more useful ie actually build temples rather than drag stone to the temple.
And of course all a horse can do is drag.Dummies can lift trash cans, scrub dishes, sweep floors, flip burgers, mow lawns, plant daffodils, prune roses, pick litter, blow leaves, hump water bottles etc etc.I don't see them disappearing any time soon.
A. Prole more os less nailed it above. Clark´s scenario has been the norm since the pharaohs yet low IQ people have continuously found jobs since then.
In the end, even if every smart person refused to hire a under 80IQ prole and used robots instead, the unemployed proles could simply vote themselves benefits equivalent to their former wages in the form of welfare and smart people would still be marginally better because robots are more productive
The unemployment of low IQ people in this case would be blamed on technology by themselves and by their usual supporters. But once welfare was removed, the whole dynamics I have described, with help of others, would play themselves.
There is a cost of adjustment for the individual. If your jobs is eliminated, you will face a time without income and the costs of relocating to other area and maybe other city. Better, from his POV, to play the welfare card
The sub-80 contingent is a minority of even the most IQ-challenged parts of US society. They aren't going to be able to vote themselves squat.
"The sub-80 contingent is a minority of even the most IQ-challenged parts of US society. They aren't going to be able to vote themselves squat."
Is this one factor that a welfare state must have a minimum average IQ? I suppose most people need to be smart enough to vote themselves benefits. In addition, I suppose the people adminstering a welfare state must also be smart.
Lots of things were true at the time of the pharaohs that eventually stopped being true due to technological advances. History is full of discontinuities. The "it will be so because it has always been so" is not a persuasive argument.
You can see by looking at rising inequality that low IQ labor is less valued.
Immigrants displacing low IQ Americans: But why doesn't the market just increase the number of jobs for low IQ people to do? If the market is so capable of making jobs for low IQ people it should be able to supply jobs for 2,3,4,5 times as many IQ 85 people as currently have jobs.
Horses could do some work better than humans. They could also tap food supplies (grasses) that human stomachs could not digest. So there were ways to feed them and get work from them that was more valuable than the work it took to raise and tend to them.
Horses got out-competed by machines. So did dogs for many purposes. Dogs also lost ground to better flooring that kept out rodents. Dogs also lost ground to electronic security systems. Substitutes jobs for horses and dogs did not pop up.
People would really prefer to use less domestic labor. If people can avoid, for example, letting a house cleaner have a key to their house they'd just as soon do that. The Roomba isn't powerful enough to replace a house cleaner. Neither is a clothes washing machine, a dryer, or a dish washer. But picture the technology of 30 years from now. Why use a human cleaner?
You have to deny the potential of robotics in order to see a future for 85 IQ labor.
Burger cooking is automatable. Floor sweeping: there are already floor sweeping robots. See my post update above.
"But why doesn't the market just increase the number of jobs for low IQ people to do? If the market is so capable of making jobs for low IQ people it should be able to supply jobs for 2,3,4,5 times as many IQ 85 people as currently have jobs."
In Sweden, people just go to "early retirement" and "sick leave" to make spots open for people. In addition "labor market political activities" soaks up some capacity in the labor market. I suppose the people in Sweden (mostly the Social Democrats) are smart enough to know that the market can't create plenty of jobs, so they use the government to create make-work jobs (such as "labor market political activities.") Sweden and Denmark are years ahead of the US and Great Britian in labor market policy.
When you look at all the problems of the world and get down to the real solution it is population control. There are too many people for the resources we have and too many people are leaving a major footprint on this earth.