So, Where Are All Those Robots?
theatlantic.com
theatlantic.com
If labor is less valuable, it doesn't have to mean fewer jobs. It doesn't even have to mean fewer hours worked. It can just mean less money going to labor (i.e. lower wages).
If our economy is operating below its productive capacity, which it is, then our levels of production are constrained by how much people can spend. And if consumers get their incomes from labor, then automation means consumers have less spending money.
If consumer incomes are constrained by wages, you'd therefore expect our level of production to decrease with increased levels of automation. And because economic policies are keeping the hours worked high, you'd expect the amount of production per man hour to decrease.
Hence...
"1. The U.S. economy is in a productivity recession."
Furthermore, if firms need to keep production levels low in order to clear markets, you'd expect that they wouldn't have much of a need to figure out ways to boost production.
Hence...
"2. Companies don’t seem to be investing in technology nearly as much as they used to."
Technology and automation allow the use of resources to be more efficient. One of those resources is labor. That means that in cases where we're actually using labor for production -- instead of, you know, just as an excuse to give people incomes -- we can get that labor really cheap.
"3. Globalization is a much bigger deal than automation for work and wages."
Globalization is automation.
But because we're Americans and we believe in the American dream, we need to force the labor market to provide incomes for hard-working Americans. Naturally, since many jobs amount to little more than token busy-work, we should then expect the labor market to become exceedingly boring and uneventful. After all, these are just fake jobs propped up by economic policy.
Hence...
"4. The U.S. economy’s creative-destruction engine is broken."
"This isn’t what the end of work was ever supposed to look like."
Yes it is. This is exactly what it's supposed to look like.
I think you may have gotten this backwards.
If you replace four pipe bending guys with pipe bending robots and highly trained maintenance mechanic, do you really think the highly trained mechanic is going to get paid less than single pipe bending guy?
Automation should increase wages for those lucky enough to get employed.
>Globalization is automation.
Globalization predates automation by at least few centuries.
Of course there's other reasons to employ people (parent mentioned spending power, I would add political influence to the list). But the labor part of the employment decreases in value.
Historically the period from inventing the assembly line to the large scale deployment of factory floor robots is roughly from 1920 to 1990. The same time period saw the highest wages/GDP ratio in US than ever before or ever since.
2) Those pipe-bending guys are now unemployed and willing to work for les thanks they did at their previous job.
If we assume that the most mundane jobs get automated first (as has happened so far) the effect of automation should again be rising wages for those who get to keep their jobs.
That seems intuitively correct, but can potentially be counteracted if (some significant share of) the jobs that aren't completely automated away are automated enough to reduce entry qualifications, such that (combined with displaced workers from fully automated jobs) there is an increased supply, relative to demand, of qualified labor, driving wages down.
Jobs that are neither automated away nor simplifies by automation such that the qualified labor supply is expanded should see higher wages, but that may be a very small share of pre-automation jobs, and may not even be the majority of post-automation jobs.
> Finally, it’s conceivable that the ostensibly tranquil and low-turbulence economy is masking something more disruptive underneath the surface. Ryan Avent, the author of The Wealth of Humans, has thought about this question deeply and offered a plausible explanation. In his telling, automation has created an abundance of labor, including machine labor and human labor. Just as rising supply typically leads to falling prices, the oversupply of labor has put a downward pressure on wages. Companies, seeing that they have access to cheap labor in a slowly growing economy, invest less in new risky technology, which leads to less productivity growth. High employment, low productivity, low wage growth, and automation can all live together in the same story.
Automation is a way of reducing costs; it is a way of making our scarce resources go farther.
On the micro level, it is a Kaldor-Hicks improvement [0]. This means that there are winners (all of society because goods are cheaper) and losers (people who have lost their jobs) - but, overall, there is a net gain. We can tax some portion of this net gain (e.g, with income tax, or corporation tax if you are really worried about foreigners) and compensate the losers (with training so they can be employed in a different industry).
In many ways, the "arguments" people have against automation are similar to those against immigration, and are befallen by the same fallacies. The difference is that we don't have to give robots health care. We already have a nice case study on rapid automation (and how the sky didn't fall in) - just look at the Luddites.
My main concern with current levels of automation is that, with patents and network effects, we are creating a new generation of monopolies (Google, Facebook, Uber, AirBnB) with very little oversight.
Or in short, work is ending, but instead of creating a post-work economy, we're creating bullshit jobs.
David Graeber did nothing wrong.
No one is claiming it is a concern for 2017, no? My understanding is that the debate is more what happens in the next 10-15 years and not about the short-term situation.
Yet every time it happens, there's a reorganisation, and eventually the people (or the kinds of people) who did those jobs are employed doing new things.
Is there some reason why everything is suddenly different this time?
Here's a couple of ways that spring to my mind, for how society could absorb the labor that would be freed up if driver automation happened:
- As increasing numbers of professional people around the world are becoming cash rich but time-poor, there is, and should continue to be, a growing demand for people to provide domestic services like cleaning, home maintenance, personal administration tasks, child-care, etc.
- As the number of machines continue to grow, there will be growing demand for people to work on the machines. Whether it's designing, programming, manufacturing, testing, transporting, fitting, servicing, cleaning, replacing, or disposing of the machines, there will be plenty of work for people of all kinds of aptitude.
We are just now seeing the impact of the introduction of computers into the workforce and that is causing all manner of issues, but the response to that so far is the best case scenario when compared to say the industrial revolutions 300 years of massive change.
No need to be a Pollyanna. Unemployment rallies didn't lead to war, they were just unpleasant.
http://www.chicagotribune.com/business/081024-great-depressi...
Economies are constantly changing and retraining and job migration are our norm, not our exception. Regardless of the current political climate, the reality is that you're not entitled to keep a job in some dying industry. Change has always been here, as much as we try to deny it. Hence the admission that we can't keep this train running and things like UBI are becoming politically feasible.
They are worried about a systematic removal of jobs at a massive rate, across the spectrum. So for instance, if you look at agriculture workers in the US it took from 1910 to ~ 2000 to get from 30% of the economy employed in farm work to ~2% (which is likely the minimum). Along the way you saw the massive labor unrest, cities had to be completely re-engineered, civil rights and family restructuring, 2 World Wars, etc.
It is a reasonable fear that something like driver-less cars, which could impact a similar number of employees, could have just as big an impact, in a shorter time frame and that might not be all for the good.
Its also possible that it was other factors that led to all that turmoil, or that we are better at societal change now, or that there is some new technology around the bend that gets unleashed due to this that prevents similar problems this time around. All of those are reasonable reasons to be optimistic (Pollyanna-ish if you will), but they don't cancel out the reasonable reasons to be pessimistic.
And the Great Depression wasn't an industrialization-related economic shift; it was more of a hiccup following the tectonic shifts in the nineteenth century part of the industrial revolution.
Again, that's not my point. It was large amounts of unemployment up to 50% in urban areas. If the GP is correct it should have led to a civil war, but it didn't.
Yeah, that's why we call it "The Great Unpleasantness", and not something like "The Great Depression".
It'll continue the "rust belting" of America, where areas that can't compete, won't, and will be dependent on aid from the government. Again, this has been going on for a long time. Just because Millenials have only noticed this recently doesn't change a thing.
This is also why some think some kind of UBI is going to be in the works in the near future. Society can handle x amount of permanently impoverished areas, but when does x become too large? The whole point of things like job tax credits, welfare, unemployment, re-training, local investment, etc is to keep people from civil unrest. UBI is just those things without the pretense that you'll ever get a job. Truck drivers in their 50s who never went to college aren't going to become competitive software developers no matter how much "training" you throw at them.
And there's been all sort of arguments (that I'm not qualified to evaluate) that the US Civil War was a direct result of economic conflict, due to the North being a much more economically productive region than the South.
When all the truckers lose their jobs they will not all suddenly be able to be project managers or accountants.
https://qz.com/879605/mercedes-made-a-crazy-van-with-built-i...
In the case of your delivery example, a person is not necessarily required to drop off the package. A system could be setup where you can request a delivery time frame for your package (e.g. between 7-9pm). When the autonomous delivery van arrives, it parks outside your house and notifies you of its arrival. You go down, scan your card and the van dispenses your package. The van then goes off to its next delivery.
this is easily automatable.
Some are. If you look at recent-ish job growth you're only seeing part-time and retail growth, or other low paying jobs that have no upward mobility. The good jobs that lead to careers, decent wages, etc have been slowly been made more and more efficient. In the past you'd see a demand for more good jobs but now with all manner of automation its just not happening as much. A IT department just 10 years ago probably needed twice the staff as one does today. AWS/Azure/Whatever cuts out a lot of middle-men, and that's just one example of how things are moving. Every industry is being affected by automation and no, its not going to look like a sterotypical 1950s sci-fi robot.
I'd hate to be a recent grad without a super-marketable skill right now.
I work in public digitization and we're still a lot of years from implementing things like self driving cars. We're running tests of course, but a lot of places the road markings, or lack there off, aren't good enough for the current tech. Heck, in Denmark we also have a lot of specialized road sings that the AIs can't read yet.
Similarly RPA and machine learning are just in the trial stages with very few human resource available to make good use of it.
What we can tell from all these tests and proof-of-concepts though, is that the technology is going to be viable sometime in the future.
No one can predict exactly what will happen of course. As I've already mentioned it'll certainly create new jobs, but those jobs will require higher educations, and I think that's the challenge which worries us in politics. Because the people that are going to be replaced by self driving cars won't be easily reschooled for something else.
https://roboticsandautomationnews.com/2015/07/21/top-8-indus...
1. Fanuc
2. Yaskawa
3. ABB
4. Kawasaki
The first one mostly has openings in Michigan, concerning machine vision and embedded NN -- from what I can tell.
Regarding point 2, can this be explained by the trend of turning CAPEX into OPEX by leasing and similar? I only heard about it in passing so maybe thats just a misunderstanding on my part.
In fact, TFP (which is what they're charting) is the mystery factor that explains why earnings grows differently than its inputs, labor and capital.
It makes no sense to me to talk about TFP when moore's law is making devices cheaper and VCs / ad revenue are subsidizing omnipresent software.
The ATMs argument is tired -- even if it's right it's 30 years old.
Foxconn replaced 60k workers on their shenzhen campus with robots. How can this article not cite that?
Weak.
The show generally covers North American manufactures where you would expect far greater automation but empirically I have observed the opposite.
For example the other day I watched a Canadian worker skillfully weave a hammock. It looked like it could have been automated but I lacked the mechanical mind to envision how.
So given that manufacturing is actually a small subset of the economy but probably the most viable for automation (ie robots) its sort of disturbing to see an almost regression particularly in North America (again my empirical observations).
Service based jobs like hospitality I would imagine to be far harder to automate (or maybe not).
I have to imagine there are big wins waiting for by have more general manufacturing equipment that can itself be mass produced.
I think this pretty much sums it up...
Wow, so capital investment slowed from 2007-2016. I wonder if anything happened during that time period that may have deflated these growth rates.
IMHO the robo jobpocalypse is as overhyped as the paperpocalypse was in the early 1980's
As the amount of businesses and transactions has increased by several orders of magnitude the overall effect is that more paper is being used than ever before.
We already know how to create paperless offices that are successful and reduce labour requirements, and yet so many businesses still aren't taking full advantage of that. That's a lot of lost jobs before we even start considering new technology replacing people.
But ask the people in the printing presses, wood mills, etc how "paper" is doing. Just because it isn't gone doesn't mean that there are headwinds for paper.
People used to say the same thing about pay phones. Cellular phones will never replace pay phones/etc. Look how quickly they disappeared.
But I agree, paper is going to be around for a while, but its usage is in decline.
Then again in a lot of office applications that is irrelevant, meaning paper is of little use to them.
sheeesh of course robototization adoption is slowing, most manufacturing is already there and until the next tech generation comes we reached the optimum split between local automated vs cheap manual outsourced