Slashdot Mirror
Robots Appear To Raise Productivity Without Causing Total Work Hours To Decline
Hallie Siegel writes: We often read about the economic impact of robots on employment, usually accompanied with the assertion that "robots steal jobs". But to date there has precious little economic analysis of the actual effects that robots are already having on employment and productivity. Georg Graetz (Professor of Economics at Uppsala University) and Guy Michaels (Professor of Economics at the London School of Economics) undertook a study (abstract) of how robots impacted productivity and employment between 1993 and 2007, and found that "industrial robots increase labor productivity, total factor productivity and wages." And while there is some evidence that they reduced the employment of low skilled workers, and, to a lesser extent, middle skilled workers, industrial robots had no significant effect on total hours worked.
This is important because it seems to contradict many of the pessimistic assertions that are presently being made about the impact of robots on jobs. What I am especially curious about is post-2007 data, however, because it's just in the past few years that we have seen a major shift in industrial robotics to incorporate collaborative robots, or co-robots. (Robots specifically designed to work alongside humans, as tools for augmenting human performance.) One might reasonably suspect that some of the negative impact of industrial robotics on low and middle skilled workers pre 2007 could be offset by the more recent and increasing use of co-bots, which are not designed to replace humans, but instead to make them more efficient.
This is important because it seems to contradict many of the pessimistic assertions that are presently being made about the impact of robots on jobs. What I am especially curious about is post-2007 data, however, because it's just in the past few years that we have seen a major shift in industrial robotics to incorporate collaborative robots, or co-robots. (Robots specifically designed to work alongside humans, as tools for augmenting human performance.) One might reasonably suspect that some of the negative impact of industrial robotics on low and middle skilled workers pre 2007 could be offset by the more recent and increasing use of co-bots, which are not designed to replace humans, but instead to make them more efficient.
APP
Mainly because work is not a set amount. We don't need X, and never need X+1. The amount of work that we want to be done so far exceeds the amount of work we need to do, or can do, that if we replace every single job in the entire world, in twenty years, all the new people will have created new jobs.
Give clothing to every single person in the world? We want more than one outfit. Give us 100 outfits each? We want to each have a unique, handsewn outfit. etc. etc. etc. Give us all sex bots and we will each want two sex bots for a threesome.
That's the nature of mankind.
No jobs? No talk to me when mankind has terraformed every planet in the solar system. Till then, stop being a ludite.
excitingthingstodo.blogspot.com
What is this, some grade school hack? If you produce more stuff with the same number of people, you've just told me that you've used robots to replace workers. Because if you can make and sell more stuff, then you can hire more people to make that stuff. Hiring robots instead of people costs humans jobs, plain and simple. I'd also be interested in seeing if "robots" included self-service terminals or whether they just count robots that move and build things.
It would be one thing if we were losing population and there weren't enough workers to fill the demand, but there are 80,000,000 more people in the world every year, and we keep making more.
Is it just my observation, or are there way too many stupid people in the world?
If total work hours aren't declining, bots, what would you say you do here?
The problem is not robots, its illegal immigrants. They steal jobs. They take jobs for less than the prevailing legal wages, get paid under the books so they avoid the taxman, and cause social strife. I'd rather robots take jobs over these illegal aliens.
Forward thinking employers - the ones who buy robots - are also more competitive. (Shock!!)
That's maybe 1% of employers.
This will change as use of robots becomes widespread.
The study concluded that productivity increased while hours worked stayed the same. As the human population grows and automation increases, it's not enough that jobs are not lost. New jobs must be created.
In the absence of robots, the higher level of production would have meant new jobs, but that is not longer the case. In effect, a job not created is a job lost.
"He took a duck in the face at 250 knots." -- William Gibson, Pattern Recognition
If robot's are making such strides in productivity, what are the flesh-n-blood drones doing to deserve the pay ? Obviously robots are doing a lot as the productivity is noticeably increased, generally means, people are getting paid the same for doing less. Or putting it in other words, the jobs that robots are performing, could be performed by less reliable, more expensive and larger number of human beings. I am wondering how soon the management will realize that they can replace most humans if not all, by robots and be done with the productivity loss, slacking on the job, etc. problems ? And What will happen then ? Will the humans take jack hammers, chain saws and attack the entities who dislocated them from their jobs and caused the loss of income ? Let's say the employers gradually phased out humans by not hiring to replace retired or otherwise lost workforce, what will happen to the unskilled labor ? Not everyone is cut out to be the programmers for those robotic systems you know...
__________
The more I know people, the more I love animals
Robots eliminate crappy, boring, mind-numbingly repetitive tasks that a human shouldn't have been forced nor encouraged to do. Sometimes those jobs are replaced with a dual job of babysitting the robots while doing some other boring task that hasn't been roboticized yet. Also, there's always the design and repair of robots.
Overall, increasing efficiency (often called "eliminating jobs") is a good thing, but can both displace workers and further concentrate profits.
Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
In the past thirty years, people with low wages haven't kept up with inflation. Also, the wealthiest of the wealthy are the main people feeling the effects because the middle class is being gutted. This has little to do with robotics though and just cheaper labor.
I'm all for robot laborers. If all the jobs are being done by robots, someone needs to discuss how people who can't get jobs survive. Maybe they can make artistic statues of the wealthy or compose songs for them.
Jobs have been automated millions of times. Every time, it's the same cycle, with on variation that happens often, but not always. The demand for goods and services is not perfectly elastic, so increased productivity for a specific task very often results in fewer people being employed at that precise task. As an example, milking machines mean that fewer people are needed to milk cows.
However, prices ARE somewhat elastic, so as the increased productivity reduces the price of milk relative to substitute goods, more milk is purchased. That increases the demand for dairy inspectors, milking techs, cheesemakers, etc. The net result is that a portion of the workforce moves from the simple job which a machine can do (and which is low paid) to the jobs which require human judgement, such as dairy inspectors.
We've been through this for every machine we have. The industrial revolution is the period in which many, many tasks formally done be humans began to be done by machine. And the standard of living improved by an order of magnitude.
For hundreds of years now we've been getting more and more machines every month. We've done, and the results are in.
The first electric programmable computer was installed in 1943, and now there pretty ubiquitous. Give robots another decade and they'll catch up.
Or you could say they're already here. I have a robot which washes and drys my dishes, another that washes my clothes, two that make me ice. Several which play video content. One which opens and closes my garage door. Heck, they're everywhere.
Is it just my observation, or are there way too many stupid people in the world?
There wasn't any special leap in robotics in 2008-2009.
There was a special leap in government spending though (remember the "stimulus"?). It turned a recession into a depression (which we pretend we don't have by simply declaring that non-workers are "no longer in the labor force", which is like saying I'm not fat because these pounds are no longer in my weight force).
There wasn't any giant leap in robotics in the 1930s either. There was that leap in government spending after a crash though ...
Every new tool or technology has ultimately enlarged the economy, and increased the number of jobs -- after causing temporary disruptions, like putting buggywhip makers out of work. For every buggywhip maker who lost his job, thousands of jobs have been created in the auto industry and other supporting industries (paving roads, transporting fuels, R&D of improved airbags, etc.).
There are more people employed today that at any earlier time in history, and most of the people who are employed today can thank some recent technology without which their job wouldn't exist.
The more disruptive the technology, the more jobs it ultimately creates. It's pure ludditism to think that robots would be the first exception to this rule.
That that is is that that that that is not is not.
The next threat is to white-collar jobs from general-purpose computers, often running AI-based algorithms, that make those human functions obsolete. My financial advisor is in danger of being replaced because he mostly recommends portfolio changes advised from the corporate back office, not his own diligence. To be fair, I get lots of additional advice beyond that in terms of other related activities that his Mother Ship does not supply.
It may be that technology increases job opportunities in the long run. However, as Keynes noted, in the long run we are all dead. The problem is to transmute people with obsolete skills (think paralegals) into employable people without the gut-wrenching economic ordeals that usually occur.
look at the jobs that are being created. They're low paying service sector jobs that are impossible to Unionize. I know, the anti-Union hate on /. is pretty strong, but there really isn't any other way to raise wages for the general populace (excluding geniuses and a few lucky /.ers who didn't see their jobs outsourced).
Basically the Manufacturing jobs from the 70s were replaced by McJobs in the 90s. You traded $70k/yr + benefits for $20k/yr without (unless you're lucky enough to live in a state with socialized medicine for the poor).
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
Robots do increase productivity. Often it opens up jobs in higher skilled areas, like QA people that check the jobs that the robots do to ensure quality. We see this a lot in the Auto industry.
The problem is what happens to the lower skilled people that get displaced by the robots? They may not have the skills, or the aptitude to learn those new skills, to do the new jobs that the robots make available. Now you have a bunch of people that used to be productive that are now unemployable.
What do we do with them? Sure, some of them might be old enough to retire. What about the person that went to work for GM right out of high school? Now they are 40 or 45 with no real skills other than what they learned on the assembly line. They probably earned a pretty good living on the assembly line. Now they are unemployed with no college degree.
Whose responsibility does it now become to support these people? The company? Not bloody likely. They put the robots in to save money. Robots don't get sick or go on maternity leave or get pensions or 401K matching. The government? Society at large? Who knows.
Bears deficate in wooded areas..
Two factories make toilet paper, one introduces robots and doubles its production and also profit margins, the number of employees stays the same. There is no impact to those employees, but the other factory goes out of business. That is where the jobs get lost and that is what the study does not measure. Same amount of toilet paper is produced at twice efficiency and half of the jobs get lost in the overall economy.
If programs would be read like poetry, most programmers would be Vogons.
The industrial revolution is the period in which many, many tasks formally done be humans began to be done by machine. And the standard of living improved by an order of magnitude.
The Industrial Revolution started what, 200 years ago? History didn't end after that. There were changes that led up to the outbreak of automation, and society has been changing continuously since then. There's no reason to believe there won't be another revolution that will again change a bunch of the rules we've so audaciously set forth.
You mention price elasticity, an economics term? Consider that that discipline was first brought into being in the 18th century. And then consider how long civilization existed before then and how many times before then people thought they had everything more or less figured out. (Well, to be fair, that whole mindset seems to have gotten a lot more prevalent from the 19th century on, but people regardless didn't perceive themselves as profoundly ignorant of the way society worked.) These understandings we've built are predicated on global markets functioning in roughly the manner they function now. Our economic models break otherwise because we've never had global markets on this scale before so we've got no alternate method birthed from a different economic milieu with which to compare our existing understanding of how employment and trade work. And that's not even getting into macroeconomics...
People have been profoundly certain of the inevitability and absoluteness of many, many things over the ages. Many times they've been wrong. It's a little premature to state with such confidence that our wage system isn't going to wind up alongside the shattered statue of Ozymandias, half-buried under the desert sands (enlarged by climate change, naturally).
We have no reason to believe that capitalism will endure until the human race ceases to exist as a species. We also have to acknowledge that if humanity will eventually stop being capitalist in the way that we understand it today, that capitalist economics contains no way to predict its own demise. And as a result, we can't use orthodox capitalist theory to prove the continued validity of the orthodox system it both describes and espouses. That's circular! The future is mired in uncertainty. We have every reason to expect that many fundamental rules of economics will be broken at some point — it's just a question of when.
If you were to poll people who once worked in bulk goods packaging you might find that they are working even more hours at their minimum wage jobs because they lost their jobs on the assembly line that barely kept their families fed. Since 2002 something like 85% of jobs in the bulk packaging world have gone. This, with a huge increase in bulk packaging output.
You really don't get it.
The invention of the automobile simultaneously made life much cushier for humans (try getting around in a horse-drawn stagecoach for a while), and created millions of jobs that previously didn't exist (from repairing brakes, to paving roads, transporting fuels, and doing R&D of improved airbag designs).
Every new tool or technology has done the same thing: increased the average standard of living, and also resulted in a net increase in the number of jobs. You don't even have to be an optimist to believe this; you just have to observe what has resulted from every other technological advance. There's no reason to believe robots will be the first exception to the rule.
We are swimming in dozens of technologies that didn't exist 100 years ago, and there are also exponentially more people employed today that there were 100 years ago. That would be quite a contradiction, if the technologies that make our lives better also caused the number of jobs "to decline." They do not. (Although they do make one's time spent working much more pleasant. Driving a tractor with an air-conditioned cab really beats wrangling a team of oxen.)
So there is no contradiction, and that is very fortunate, because how would seven billion humans earn a living if not for all the technologies that created exponentially more jobs? Do you say "please stop with" every other technology that has lifted people out of poverty and caused the number of human jobs to increase? (No, you do not... in fact you use a piece of technology that's extremely advanced, by the standards of 25 years ago, every time you post to Slashdot.) So you and your ilk better not say "please stop with the robots." That would put a hard cap on the number of jobs, and ensure that average working conditions won't improve, and (if world population continues to increase) doom us to ever-increasing unemployment rates.
That that is is that that that that is not is not.
Oh I'm no Luddite and hardly a peasant. :) I love tech. Nowhere did I suggest limiting technology. What I said was that people have to support themselves and have some decent kind of life. If robots take the jobs and you don't redistribute wealth (which is not the same thing as socialism), there will be a revolution, regardless of the carrying capacity of the West. We don't have real famines yet because most people can still find enough paying work to buy food, or they rely on charity. But you can't humiliate the majority of the population, just so the rich can have the undiluted spoils of capitalism.
Hmmm. Now that you mention it, perhaps I am beginning to feel a bit socialist. And so will the unemployed masses in a robot filled world. Adopt a Star Trek economic system (post scarcity) BEFORE the pitch forks come out.
"He took a duck in the face at 250 knots." -- William Gibson, Pattern Recognition
Nobody has ever said that humans can't contribute in workplaces run mainly by robots. The question is can capitalism adapt its model to deal with this scenario.
In other words, will the companies using the robots realise they still have to pay employees in order to produce customers?
How do you write an entire article about productivity, hours worked, and employment without a single number being used? The authors describe their methodology and their conclusions but none of the data that was analyzed to reach their conclusions.
Manipulate the numbers the right way and you can get absurd conclusions like this article.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
What you are saying is that you're willing to sweep the displaced under the rug and forget that they exist - since there's something new around the corner.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
You want automation? Fine - have a plan for integrating the displaced in some snark-free way. Otherwise, enjoy being on the business end of a malfunctioning ED-209.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
It essentially allows the same worker to do more per hour. However, unless somebody actually purchases the output, the factory is limited to the amount of extra widgets it can actually sell.
The bottleneck in the cyber-age economy is consumers, so far. The same or fewer workers can produce more, meaning the proportion of jobs that increase to absorb the extra products are not there to match the output increase.
Nobody has figured out how to get more and bigger spending-consumers. Most of the revenue and profits are log-jammed at the 1%, who don't need 500 iPhones each.
Taxing the rich seems the only known way to free the revenue and profits to flow back into the middle- and lower-class consumer. If you have a another way to balance that part of the system of economic flow, I'm all ears.
Table-ized A.I.
Aside from very few lines of work, provide some certainty by training the people already in existence. Another option is to adjust compensation to get the people you want.
There is no skills gap, just pay and training gaps.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
I put the hyphen in the wrong place. It should be "more and bigger-spending consumers".
Some smart-aleck would make a joke about already having enough fat consumers.
Table-ized A.I.
Ultimately what will be (I believe) the best solution is some form of tax on commercial automation, that will be used to fund a Basic Income. As automation increases and replaces more jobs, the fund will increase and can support more people. The stock answer to automation is "but that creates jobs for robot engineers/repairmen", but eventually, machines will be able to repair one another, and engineer new designs. Unfortunately, the transition to a fully-automated economy will be slow enough that we can't suddenly drop into a "everything free for everyone!" economy, so a transitory solution is required.
The detail devil is that the tax will need to be well-calibrated, so that utilizing automation is still cost-effective, and that robotics startups will be able to get off the ground. Another issue is what exactly counts as 'automation': do more-efficient tools like carbon-fiber handsaws count as 'automation', since they're modern tools which increase efficiency, or must it have moving mechanical parts like a chainsaw? What about a spinning saw rotated by someone pedaling on a stationary bicycle?
Perhaps that issue could by avoided by setting a standard of productivity per man-hour used, with any excess (through overwork, or better tools, or more usage of machines) being taxed. The rate of 'base productivity' would have to be set by the government, although I'm unsure how that would be calibrated, particularly for a new fully-automated factory that never had human workers. It should also be predictable such that a startup would know how much they would be taxed, when determining if the venture would be profitable. The downside of THIS method is that it's difficult to assess productivity in some fields (computer programmers and similar), and many jobs which utilize automation are service jobs; the field of medicine changes so frequently that assessing 'productivity' there would be difficult. A strict definition of 'productivity' as 'billable value' might help take care of service jobs, but assessing productivity of the more abstract cognitive jobs remains elusive. Call-centers that use automation to provide free tech support for a purchased product, for example, would have their automation taxed how? A combination of the two methods may be required.
Many Americans who are unemployed, or criminals, or on welfare, already live off of $5k or less per year, so a $5k basic income would be 'liveable' to them; if you live in a house/apartment with a few other people like you, you can cover the bills. As the basic income rises to $10k, the unemployed would feel less desperate, and crime would no longer seem like a necessary evil to stay alive. For reference, paying $10k to the lower 50% of adults in the USA would cost $1.2trillion per year, around what we were spending in Iraq each year we were there; and it could replace welfare programs, so the net amount wouldn't be as much.
Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
Two words -- Hawthorne Effect https://en.wikipedia.org/wiki/... Also doesn't solve the problem that most things made in factories need to get bought by middle class people, of which there are increasingly few with increasingly diminished purchasing power.
Manufacturing jobs have dropped every year since robots were introduced while productivity has risen.
http://cdn.theatlanticcities.c...
http://www.technologyreview.co...
They've been replaced with terrible low paying service jobs.
Wages have been stagnant for 80% of those who have jobs since shortly after robots were introduced. Robots are not the only cause- but they sure didn't help.
While the unemployment rate is finally tightening up some- that's because so many have completely left the work force. Participation of working age citizens age 16 to 67 has dropped continuously for the last 14 years.
http://www.washingtonpost.com/...
Who paid for this article? The robot manufacturing companies?
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
The industrial revolution started at least over 250 years ago, fyi.
"And while there is some evidence that they reduced the employment of low skilled workers, and, to a lesser extent, middle skilled workers, industrial robots had no significant effect on total hours worked"
There is a chance that overall the numbers of hours stayed the same but as indicated here the reduced employment on low skilled worker is the problem. Up to recently there was always a way for such to go into a different branch with low requirement. It isn't possible anymore. And this is the huge problem the study misses. And this is the doom and gloom issue that not only some people on slashdot but on economics forums foresee. Furthermore the study does not seem to make a difference between part time hours and full hours as it aggregate the overall time.
C. Sagan : A demon haunted world:
http://www.amazon.com/gp/product/0345409469/
visit randi.org
You mean machines taking over physical labor has worked in the past. Now, machines are taking over intellectual pursuits. What jobs are supposed to be generated out of that?
The problem is that you have assumed an open system. The rest of the system found other things for people to do. When machines start taking over intellectual pursuits as well as mechanical, you've pretty much covered all human endeavors. The system is now closed. You need to find some way of enlarging the system and you'd better do it fairly quickly and constantly, humans have ways of multiplying and turning on each other when they have idle hands.
The only thing on that show is how crap is made. It's all pathetically low-grade tech, building artsy-crafty dreck most of us wouldn't be caught dead owning. It will never show truly interesting stuff like jet turbine construction, motherboard manufacture or the like, because the production processes of such cool things are truly a corporations' 'family jewels'.
However, "expert systems" are now removing the need for much human involvement in many jobs that used to require human perception, interpretation, judgement and decision-making.
"Robots Appear To Raise Productivity Without Causing Total Work Hours To Decline"
That's the title of the article. What isn't discussed is what are the quality of those hours, i.e., how much are people making per hour with robots. According to everything I've read, wages have been stagnant. So that's one issue I have with robots.
Secondly, we can infere that without a robot, hours worked would increase. Not necessarily a bad thing since that would probably mean that someone had to be hired to work that job. There's a large group of people who are never counted in the unemployment numbers who have given up looking for a job. These might be helped by not using robots.
Now, if my employer wants to give me a robot and raise my wages, I'm all for that.
and I'll say it again - technology INCREASES jobs, never decreases it - over the long term. Over the short term it can make certain skills worthless, putting some people out of work, but that's it.
If your position is correct, the number of jobs in Agriculture has increased over the long term.
So, for instance, the number of people working on farms has increased over the last century or so.
Yes?
That would only be true if he had said that technology increases jobs in every field—or, perhaps more pertinently, increases jobs in proportion to their current distribution.
He didn't. He just said that it increased jobs overall—that is, if there were 950 farm jobs and 50 office jobs before a particular technological advance, maybe there are 1450 office jobs and 50 farm jobs after. Significant increase in total jobs, even though people who can only do farm work got the shaft.
Now, perhaps his point could be debatable, but it doesn't mean anything remotely like what you've said here.
Dan Aris
Fun. Free. Online. RPG. BattleMaster.
"Depending on how you want to think about it, it was funny or inevitable or symbolic that the robotic takeover did not start at MIT, NASA, Microsoft or Ford. It started at a Burger-G restaurant in Cary, NC on May 17. It seemed like such a simple thing at the time, but May 17 marked a pivotal moment in human history."
So begins the short story Manna, by Marshall Brain (full text on the web). It's the creepiest dystopian story I've read in a long while... made even creepier by the breezy and cheerful way its central character (who reminds me of Philip J. Fry from Futurama for some reason) tells how the robotic revolution will really go down. Take a few minutes to check it our. Your eyeballs will pop out on springs.
As I write this I'm thinking, should I even mention this story? Someone may think it's a great idea. I'm also thinking, perhaps if enough people read the story they'll have time to think about it and perhaps find a way to stop it from playing out to the end that is described. Some appropriate response that falls short of going full frontal Luddite.
Never mind those cute robots unable to walk up stairs, though some day they will. Forget that silly stuff about Skynet, it doesn't want to hurt you, though some day it will. But the first robots may actually be... people. Starting tomorrow.
Also by the author,
Robotic Nation
The Second Intelligent Species: How Humans Will Become as Irrelevant as Cockroaches
<blink>down the rabbit hole</blink>
Mmm, no. One lesson from history would be the technological advance that is the analog computer, and the fact that this technology was lost for some number of centuries. In particular, the Antikythera mechanism (~205 BCE?) and subsequent reinvention of the analog computer (~14th century CE). Thus, technological advance is an insufficient guide, as it did nothing to prevent the Greek and Roman civilizations (and indeed every past civilization, ever) from faceplanting. Such faceplants probably had some impact on the jobs market--mead-maker for local warlord, assuming one survives?
Perhaps optimists would know these cycles of history if they were not so busy fitting straight lines to semi-log graph paper and calling things good?
I just googled reasons for the fall of the Roman empire, and while many historians have proffered many theories, none of them cited "because they employed new technologies" as a reason.
Indeed, Rome certainly would have fallen much earlier if not for their great technologies -- such as aqueducts that brought copious amounts of water to their cities, and allowed for half-decent sanitation. If you disagree, I assume you will be spending your afternoon ripping out your indoor plumbing?
That that is is that that that that is not is not.
The economy has been based on endless expansion. That is not sustainable. So we can just stop here, because your entire comment is predicated upon this idea, and it is a foolish one.
Wow... "the size of the economy can never exceed X" is right up there with
The "640K ought to be enough for anybody" statement looks like one of the most dogmatic, short-sighted comments ever, a verbal blunder perhaps topped only by Digital Equipment Corp. founder Ken Olsen's 1977 quip, "There is no reason for any individual to have a computer in his home."
Exponential economic growth is not sustainable if you limit yourself to the resources in a closed system. But if you bring in new resources from outside the closed system, the game totally changes. (You can disagree quantitatively with his estimate of "Gross World Product increasing by a factor of 10," but hopefully you agree qualitatively.) Incidentally, robots would definitely be needed to lay down thousands of square miles of PV cells in an environment deadly to humans.
That that is is that that that that is not is not.