It took several years, but Paul Krugman finally wrote a column I disagree with. There's plenty of studies showing that automation is a real threat to jobs in the future.
108
UBI is simply idiotic. Makes me think of a little vignette in a Douglas Adams book. Group of (relatively useless) middle managers crash lands on a planet. Things are bad, people feel terrible about their living conditions.
So the newly elected President declares that leaves are the new currency. The people are very happy. Everyone is suddenly a millionaire!!!
They soon, of course, realize that it pretty much takes an entire deciduous forest to buy a potato.
The current US baseline is $0 ... i.e., if a person does nothing, they have no money at all (except for government aid.) If that baseline is raised to $1,000, you don't think markets will react? That prices of pretty much everything won't go up at a commensurate level?
This is macroeconomics 101 folks (not even the college level - the high school level.) If there is suddenly way more money chasing the same amount of goods and services, the price of those goods and services will jump.
Day One of a UBI would have a lot of people happy about their leaves. Day Two would be the Fed having frantic meetings trying to figure out exactly how high they'd need to crank up the prime to stop the country from spinning into hyperinflation worse than even Venezuela.
There are NO easy, simplistic answers to wickedly complex problems.
6
Krugman is about as tone deaf as one can be on this. Thousands of workers are being replaced at supermarkets by self checkouts.Thousands are being replaced in fast food outlets by touch screens. Receptionists and security guards are being replaced by cameras and droids. And we are not even talking about about driverless buses, trains, and transports which are starting to appear. So tell us Paul, where do these people find new jobs?
17
This entire piece ignores the degradation of working conditions in the US.
Most of us have become much more expendable and thus much more subject to the throws of economic tumult which seem to be inevitable given the gutting of already inadequate regulations put in place after the last recession.
We are getting paid less (small uptick in wages does not make up for larger increases in costs of living) for worse working conditions and it seems like the solutions proposed by Warren are just iterations of promises I've been hearing for the last decade that will never get through congress.
You say yourself that Bernie's plan probably isn't workable.
You claim that a yearly, direct, trillion+ dollar stimulus to the lower and middle classes is "inadequate" then go on to claim that an infrastructure project is somehow going to be the answer to our economic woes? This can't be serious. We need to work on our infrastructure but I don't see how that is going to a better solution to our economic problems than UBI.
I have read Krugman's column for a long time and I know he likely realizes how weak the arguments he laid out in this column are.
This is a sad week for NYT. This paper, opinion and news, is obviously doing to Yang what it did to Bernie in 2016. I thought Clinton would have made a better president than Bernie but the biased coverage in this paper was egregious. I lost a lot of respect for NYT in 2016 and I am hoping to gain it back... Don't let America down
15
Krugman very sell may be right here about automation. While it may not be an immediate threat, replacement of skilled employees by Capital intensive machines is a potential problem which needs to be solved eventually. Yang gets that we need a large contingent of paId workers to consume goods, without it Capitalism cannot make a profit, withers and dies. Putting a floor below poverty by a national income may be a good idea. It would be nice to have a rational discussion, rather than a 30 second dismissal.
10
"By 2005 or so, it will become clear that the Internet'e impact on the economy as been no greater than the fax machine's." Paul Krugman, 1998.
14
"You might say that this time is different, because the pace of technological change is so much faster."
This time is different, but not because it is so much faster. It is because the part that is being automated is the human brain, not the human muscle. The industrial revolution ushered in the replacement of human labor with machine labor because the machines were faster, stronger, and more accurate. It did not render humans obsolete, because machines could not replace human brains. The digital age began with machines that could only perform 4-function arithmetic, a mental task that most 3rd grade students can perform. We are now at a point where machines have become completely dominant on a range of complex mental and physical tasks, and the list is growing every day. Driving a vehicle is a good example of a task which is not that mentally complicated, but requires image processing (vision) to perform correctly, a complicated process to replicate in a machine because of the amount of computing power required. Chess players worldwide have already been rendered obsolete by a machine. I am not sure what the political answer is, but we humans already lost the physical strength, speed, and accuracy race to the machines. If we lose the mental capability race, what do we have left to offer a capitalist society other than entertainment?
17
Discussing the societal implications to automation in our labor economy certainly belongs on a presidential debate stage. In the same way it's the center piece to another thoughtful op-ed from Mr. Krugman. But, we're assigning too great an authority on a narrow-sighted prognostication to not consider seriously that this time is different. Naturally, we all confidently suspect that new opportunities in different or changing industries will arise as portals for productivity. And a well functioning public sector could function with the kind of focus to promote career advancement. Krugman's strongest point is recognizing the inadequacy of private spending. Creating an attractive market for private enterprise to invest in human effort and engagement will require a great deal more refinement than simply fixating on automation.
In 10 years we went from almost no one having cell phones to almost everyone owning and using them. That is the pace of technological change when it works reasonably well and has a compelling benefit. So I think automation will happen more rapidly than it did in the past.
But it seems like it's getting disproportional hype. It won't be everywhere, wont' do everything, and won't be without flaws, some serious. Many skilled jobs tat require human judgement and interaction that no machine will be able to do any time soon (think medicine, teaching, technical professionals, etc.).
Beware of either/or-ism. Our brains are wired to think in binary terms, but ‘this’ does not always necessarily invalidate ‘that’.
It seems clear-as-day to me that the losers of global trade and the losers of the tech economy are one and the same—factory workers whose middle-class incomes were transfused to upper management when their jobs were outsourced to the developing economies of China, Mexico and India, and whose current minimum wage jobs are being automated out from under them as we speak. But pay no mind to the rage of white males succumbing to deaths of despair via opioids and alcohol abuse, that’s just their “white privilege” talking.
Enough!
The data Krugman cites as minimizing the impact of automation—the BrookingsMetro Automation-AI Report—is actually quite dire if you consider “some people” as human beings with families and dreams rather than merely cogs in a machine. Who we should pay no mind to are cynical apologists for the rot in our system (based on the myth that the anomalous economic boom of the post-war years would continue indefinitely), of whom the author of this column has been in the forefront for decades. Accountability is coming, and it is already grim.
2
Krugman is right about this but he is part of the problem.
The problem is not the only robots but left and right tail (in the political distribution curve) radicalism. The "left wing nuts" are not a substitute for the "right wing nuts" of Trumpism, that the American people will accept. The robot problem is just a manifestation of left wing which ironically Paul Krugman represents.Trump's greatest political strength are the scary people who would take his place.
“...one of the moderators, asked a bad question, and the debaters by and large — with the perhaps surprising exception of Bernie Sanders — gave pretty bad answers.”
I’m not surprised that Sanders gave the best answer. Sanders doesn’t live in the Fairytale world of the corporate/media, wherein for instance, Krugman and others applaud low unemployment and job creation stats while disregarding the quality of those jobs, or ignore the fact that health insurance rates HAVEN’T soared over the past few years, making health care unaffordable for most.
Sanders represents the views of the non-elite, and that will ALWAYS result in a more accurate view of reality.
4
Astonishing misunderstanding about robotics from a usually insightful pundit. It does not require technological determinism to explain the profound effects that technoscientists, business executives, consumers, and others have wrought on future generations. Think synthetic chemicals, automobility, and war machinery for a start.
On none of those or other technosocial transformations has there been anything like an informed set of democratic decisions. Robotics and associated phenomena are repeating the hope —without a fallback strategy — that the benefits to some will outweigh the costs to others. And that the civilization will be one that our grandchildren find livable.
5
There is an article in last week's Economist illuminating another aspect of Krugman's analysis: AI in general and self-driving vehicles in particular. All participants are pushing back the timelines for self-driving vehicles, which have the potential to put all truck and cab drivers out of business. But last I heard there was a nationwide shortage of 60,000 long haul truck drivers. Yes, AI and self-driving vehicles are coming, but it's not the most pressing problem this week. An enduring problem is how to retrain those who lost jobs to off-shoring and natural gas replacing coal before more become addicted to opioids.
4
It seems like linear thinking, automation will take most of the jobs aways someday and we need to plan for it. Maybe a guaranteed income is a solution. Once we get true AI most of the jobs will be gone as the machines will outperform humans by a huge margin.
1
I agree that putting the unemployed on a monthly dole is admitting defeat, an odd way to kick off a campaign.
But we may still have a chance to save some jobs from automation or at least get people trained to operate the computers. On the other hand, there is absolutely nothing we can do to make our workers a viable economic alternative to foreign workers who earn next to nothing.
So if "free" trade is the real problem, not automation, we're sunk.
2
True enough but this was not the magic of the marketplace in action. Throughout the 20th. century in the U.S. states and the federal government implemented policies that had the effect of adjusting the labor force to needs of the economy.
Mechanization automation both diminished the need for labor and provided the wealth necessary to allow more and more people to spend time not working. First children were removed from the labor force through elimination of child labor and rising requirement minimum school attendance. Then labor rights protections allowed for the 8 hour day, and other benefits reducing the total hours in a normal work-year. Social Security, union pensions and Medicare all made possible a shortening of the work-lifetime. The GI Bill and subsidies to higher education allowed millions to defer entry into the labor force for 4 or more years. Then of course there the establishment and maintenance of an enormous standing army. And, lets not forget, incarcerating millions of potential workers.
If people still started working at 10, worked 60 hour weeks and labored in their traces till they physically couldn't handle it anymore and lived out a short retirement in poverty we would have staggering unemployment and under-employment. It is no coincidence that wages stagnated in the 1970s just at the time the government stopped removing people from the labor force.
The impact of A.I. and automation demands a public policy response because it always has.
6
If automation in manufacturing can accomplish repetitive tasks more efficiently than people, what are we supposed to do? Go backwards? Ignore the competitive advantages of technology by retreating into an outmoded past? That's a sure-fire recipe for obsolescence (and bankruptcy). Is it better that our companies simply fail altogether because they can't compete strategically in the world marketplace?
Jobs aren't "lost" to automation. Some inevitably become irrelevant as innovation finds new ways of adapting. If we built cars or skyscrapers or telephones the way we did 30 years ago, those industries would be dead in America now. The jobs backward-looking politicians falsely claim they're fighting to "save" are simply outmoded. The new jobs are in the future, not the past.
2
I rarely disagree with you. But to call automation a "pseudo issue" and dismiss it so out of hand is unusually cavelier for you, Mr. Krugman. It may not quite be the key number #1 issue that Yang purports, but it is hardly a pseudo issue. You say we need greater private spending, and who can disagree. Stockpiling wealth by the financially elite is killing the rest of us. But to characterize this "pseudo issue" as "blaming robots" is beneath you. Have you considered that BOTH Yang and Warren/Sanders may be right, but simply in varying degrees? I wonder if you have actually listened to Yang make his case. It is a compelling one to me, and at least should not be simply discarded. I wonder if you really believe that in the next 20 years we will see the private spending that you prescribe. Is it even a reasonable assumption? I suggest giving that $1000 a month to the middle and lower classes, which in spirit is what Yang suggests, and hedge that bet on private spending. In the meantime, do I hear you endorsing Warren/Sanders? I think I am hearing that.
10
Debate like this is not going to take vote away from Trump.
Trump got many Americans fired up by asserting immigrants and Chinese stealing their jobs.
Now democrats are saying robots taking away their jobs? Do we really expect the voters to get excited and begin to vote for Democrats because they now hate the robots so much?
I agree with Prof Krugman. We should avoid the robot rabbit hole.
3
I wait in my favorite cashier's line. She doesn't know I love her. My wife does but just rolls her eyes. She looks at me, doesn't smile, says her required "Find everything?". I say, "Sure", and give her my joke of the day. I think her favorite was "What did the number 0 say to the number 8?" "What." "Nice belt". She has a terrific smile. She says "thanks" when through. She doesn't tell me to have a nice day. Recently she made some of her hair blue.
2
Change is inevitable and one of our greatest failures is that we have not changed our educational system and society to be in step with all the technological advances. We've also been crippled by climate deniers and simply by those that wish to profit now leaving the clean up and problems for future generations. Oil companies knew decades ago of the now impending ice melt, climate extremes, etc yet they chose to lie or hide the information. This is criminal. So in some respect yes we can't just go down the "automation rabbit hole" but the alternative to ignore what is around the corner or here now is far worse. Yang at least has some great ideas based on reality, science and basic solutions. He may not win but we need his voice in our government over climate deniers and deregulation. I don't hear anyone in current government talking about any of this because our current administration is so problem filled. We can't even agree that Trump is a bad President?!
2
What do we do with people that cannot produce enough value from their labor...?
With AI automation eliminating many jobs...eEveryone cannot be "trained" for jobs paying a living wage. Some simply will not have the capacity.
The current plan appears to be build more prisons and gated communities to keep such low-value people at bay.
I hope our leadership can figure out a better way.
3
AMEN - saying alarming things about automation is the moral equivalent of saying alarming things about immigrants.
1
Some jobs—long-haul truck driving, I’m looking at you—are probably going bye-bye. And that’s scary, if that’s what you do.
But employment, that’s not disappearing. Because computers are labor-increasing devices, for one thing. The question is, will coal miners and such get off their backsides and change, or keep dragging on everybody else?
By the way, there’s an assumption buried in Krigman’s argument that I wonder about. Why’s work so nifty, anyway? Just to keep the proles busy?
I wonder if we don’t have an equal crisis in the growing recognition that too may have jobs that aren’t really worth doing. Kinda hard to take the pride in your work (which is spozed to be a primary reason FOR work) stocking shelves at the Dollar Store.
1
Krugman sounds as up-to-date on things as Biden and his record player. My 23 year old heard this discussion during the debate and said Warren sounded clueless as to what AI is and will do.
9
Let me supply some basic numbers
Such as the relation of robot density to manufacturing employment
Number of Industrial Robots per 10 000 employees in Manufacturing:
US 189
Germany 309
1, Germany has a higher robot density than the US but a flourishing manufacturing sector. So it is NOT robot density
2. The crucial difference is that Germany outsources fewer manufacturing jobs . Partly it is cultural, partly it is the fact that by law the governing board of large companies must be made up half of labour representatives.
3. Mr. Krugman omits outsourcing as a reason for the shrinking US manufacturing sector. From the NY TIMES 1993
"Paul Krugman, .... said that for the United States, the agreement is "economically trivial." He sums up the war of words this way: "The anti-Nafta people are telling malicious whoppers. The pro-Nafta side is telling little white lies."
After those who saw their jobs move to low wage countries wrote him critical letters, he responded in 1997
"I guess I should have expected that comment would generate letters along the lines of, “Well, if you lose your comfortable position as an American professor you can always find another job–as long as you are 12 years old and willing to work for 40 cents an hour.”
Such moral outrage is common among the opponents of globalization–of the transfer of technology and capital from high-wage to low-wage countries and the resulting growth of labor-intensive Third World export
2
Note for all those proclaiming how we can replace the lost jobs with service jobs such has nursing home attendant or child care. Did you ever notice that service rarely pay enough to keep someone in a roach infested studio apartment in a neighborhood no sane would live in ? much less pay for enough Ramen to eat for the month?
2nd note will forgive you taking away their lives before they will forgive you taking away their self respect. And make work jobs, Jobs with no hope, no future, & No dignity, Which is all so many are left or are going to be left. For themselves or their children, And the intelligentsia can't figure out why Trump got elected? Hint he's not aberration his a harbinger of things to come. Has automation & Globalization strip more and more people of hope & Dignity? Well scorched earth is more attractive then the status quo. And the losers are turning to scorched earth no reason not to.
5
Wrong I am 66 I love self check out. Some should not use it, for me it is great.
I don't really take Andrew Yang or Tom Steyer seriously as candidates. We already have a rich guy with no political experience in the White House and it's been a disaster. One is enough.
1
If we had a strong social safety net with good unemployment benefits (not the kind where you have to urinate in to a cup, or prove that you are sufficiently suffering), good government-paid health care (paid by us through taxes), etc. one would not be sweating out the 'robot revolution', at least job-wise. Building a strong social safety net, abolishing the Electoral College and electing Democratic super-majorities who will restore sanity to the tax system will save America.
2
"Let me also give a shout-out to Joe Biden, who echoed Yang’s talk about a “fourth industrial revolution.”
Wish you had expanded on that. Yes, I think Biden and Klobuchar are the only people who are rationale and can win in our Electoral College.
Can you remind your fellow columnists that telling us who would win in a poll with Donald Trump is WORTHLESS, unless one corrects this for our Electoral College???
I think Warren and Sanders will especially do TERRIBLE in the Electoral College.
Thank you.
In 1979 I was in Egypt doing a public transportation study for USAID which entailed interviewing managers at all levels. I was surprised to discover that even managers at the lowest levels had personal servants fetching coffee, etc. I later learned that these persons subsisted on only a few dollars a week. When I mentioned this to a colleague who was something of an armchair Marxist, he said something like. "Well unlike the US, at least everyone has a job."
1
Interesting that you mention Vonnegut, I am more convinced every day that when Kurt arrived in heaven he asked God to be his lead scriptwriter. Only KV's mind could come up with the reality we're living through.
Yesterday as I walked down the street in the financial district of Chicago, I overheard a conversation of two men in suits behind me. One said to the other, "We don't rely on people, we rely on automation to get the work done." Honest to goodness.
I don't know what work they referred to, but the first thing I though of was Yang.
2
The primary supermarket where I shop has only friendly human checkers, baggers, a very clean store, help locating and retrieving items and many other conveniences. Some in the area complain about the prices which are not much above other stores. They seem to never lack help because they pay well and have great benefits. They've expanded from one store in Lakeland to much of the South.
1
My Costco just put in some self-checks and is really pushing customers to use them.
I have shopped there close to 20 yrs and have seen cashiers and baggers move up the ladder to better paying jobs there.
So if there are no longer cashiers, there will no longer be entry level jobs for people starting out. Sad.
2
Bill McKibben, in his excellent new book, "Falter," talks about AI and automation, and concludes that it will most certainly make economic disparity/inequality even worse than it is now, as the owners of capital machines and AI services reap more concentrated benefits for themselves at the expense of everyone else. This is not like the advent of other technologies, this is different, and more dangerous to the overall masses.
4
Get a cashier!
Did Krugman publish this on a fax machine?
3
Bravo! The NYT is so lucky to have Krugman and Leonhardt.
Machines taking over employment through 2030: What’s the story? What sorts of things do we need to do? See http://worksnewage.blogspot.com/2019/07/the-next-11-years-in-united-states-jobs.html.
"Which makes you wonder what Andrew Yang is talking about. Yang has based his whole campaign on the premise that automation is destroying jobs en masse and that the answer is to give everyone a stipend — one that would fall far short of what decent jobs pay. As far as I can tell, he’s offering an inadequate solution to an imaginary problem, which is in a way kind of impressive."
Thank you, Dr. Krugman. I've spent a few minutes puzzling over what Mr. Yang has been saying and, as always, you've explained it. "An inadequate solution to an imaginary problem."
1
Andrew Yang may be a footnote in this election but his concerns are real albeit a bit ahead of its time. Yang rightly states that the impact of the digital age and AI will be of increasing concern in part because its spread is less observable and apt to elicit less of an emotion response than another human replacing you. Robots for the Dems are like immigrants to Republicans.
The problem of robots can wait. What we need immediately is a new government that truly has the people and our Constitution as its primary concern. Secondly, we need new governance structures in place to re-build and strengthen our democracy so it can not be as easily hijacked in the future.
To get this done, we need to ban 'free', 'give', and 'guarantee' from the Democratic Party's vernacular.
3
Like many of those commenting here much of what I understand about economics has been as a result of reading Paul Krugman's New York Times pieces over the years. Nevertheless, I believe that the UBI would result in the transformation of our society for the better.
Sadly, I believe that Yang has mistakenly chosen to stress the UBI as the only effective means of coping with the future effects of automation on employment. I think Yang uses such language because it is helps him create a politically compelling (albeit highly debatable) narrative.
Nevertheless, it is possible to be agnostic about automation's effects on aggregate employment while still embracing the UBI as a progressive path forward. UBI makes sense as a policy prescription because it would in large part end poverty and precarity in America.
Critics of the UBI will contend that the UBI is insufficient to eliminate poverty and they are correct. The dollar value of existing transfers for many families exceeds the thousand dollars a month that Yang's Freedom dividend pays. But progressive UBI supporters understand and support maintaining most existing safety net transfers.
4
As for the UBI's effect on ameliorating the precariousness that is increasingly endemic among America's poor, working, and middle class - there is little doubt that 1 to 2 thousand dollars a month would provide families with a much needed cushion as they navigate a new world of work that will require folks to train for and transition into and out of multiple employment situations. The boon a UBI would represent for families caring for their kids and/or elderly parents should be readily apparent as well.
As to whether or not the UBI is practicable there may be reasons for doubt. For those of us on the left who support it, however, the UBI remains a humane and pragmatic response to a multiplicity of today's socio-economic challenges. Yang has brought the UBI out of the political dessert where it had languished too long and for that he will have my support for as long as his candidacy lasts -regardless of the robots.
3
@Richard Keefer Zero chance it would end poverty. Prices will just go up. Bet on it.
It’s true there’s a lot of work to do in America. But it is not the sort of work that built the middle class of past decades. The manufacturing jobs that could comfortably support a 1960s high school graduate have been mostly automated away. Ditto for most repetitive jobs—typist, telephone information line operator, and all the jobs that used to beat the other end of a robocall. Not to mention road map jobs lost to GPS, ETC, ETC. As for trade deals, some studies suggest that their impact has been more positive than negative. The Council of Foreign Relations reports that, in the view of economist Gordon Hanson, “ NAFTA helped the U.S. auto sector compete with China. By contributing to the development of cross-border supply chains, NAFTA lowered costs, increased productivity, and improved U.S. competitiveness. This meant shedding some jobs in the United States as positions moved to Mexico, he argues, but without the pact, even more would have otherwise been lost.” So, both Yang and Bernie are right. And the solution, if we don’t want to tear up the world’s supply chains, is to train workers to profit in the age of the globalized machine.
1
Perhaps an analogy would help. Consider the cell phone. It is a technological marvel, and its development has undoubtedly had an impact on the land line industry, not to mention the camera industry. However, there is now a moderately large new industry: inventing and producing cell phone accessories: cases, screen protectors, chargers, camera lens extensions, tripods, and many other kinds of hardware, and that's not even to mention the now huge industry of software development and maintenance that would never have existed without cell phones.
As robotic production produces old products, humans will use that as a platform to design and implement new products. And so it will go.
2
The immediate capitalistic reflex to pick the lowest overhead policy only takes into account corporate profits and shareholder value. As Warren has proposed, insuring that workers and community leaders are on boards of directors of all publicly held corporations should be the appropriate remedy. All this must take into account the intrinsic value corporations must have to the social fabric in a capitalist democracy. The technological challenges will always be there, and the need lifelong education.
2
Bravo, Paul Krugman. Thank you for calling out this relentless Chicken Little myth the has been plaguing American politics since the Great Depression. Throughout the 1930s, automation was largely used as an excuse for the crash. Recall the assembly line was a type of automation.
And thank you for pointing out that in the 1950s, production was more automated than ever, and the American economy could not add jobs fast enough. Yang completely ignores the fact, historically, automation has gone hand-in-hand with job GROWTH. Not job destruction. If you want to know Andrew Yang's vision for America's future, watch a movie called "Ready Player One."
(Tip: Don't vote for Andrew Yang.)
1
I really appreciate Krugman’s analysis here and while I am compelled by Yang’s intentions and purpose, I take Krugman’s point that it is to some extent misguided. It’s true that we have been afraid of automation for ages, and yet our fears seem to have been a bit more extreme than reality called for. However, I think it’s also the case that middle class jobs (those allowing a person to pay off student loan debt and raise a family) do seem harder to come by than they once were, in part due to automation (and also trade policy). However, what I think is more pressing is that automation, coupled with economic policy, allows tech companies to accumulate excessive wealth and power. As Krugman points out, those companies are not spending to a degree and in a way that creates economic opportunities for low and middle income Americans, or that does anything but expand the wealth and income gap. I think Yang’s proposed policies attempt to address this by forcing the hand of tech companies to invest in American curiosity, creativity, and passion via a UBI. Whether it’s the right approach is questionable, but I don’t think it’s as kooky as Krugman makes it out to be.
2
Forget the Truck drivers - think Government services like the DMV.
And these are highly unionized jobs that support underfunded pensions.
The Government Unions will be the Luddites breaking the mills.
i recently heard a presentation that was very optimistic about the future of work. One take away is that a survey of 2000 CEO's asking what attributes that are in most need in their workforce were attributes or skill not easily replicated by machines. Robots aren't very creative. Plus coal miners don't need to learn to be coders, but if there was investment in restoring depleted eco systems that would be very rewarding and safer work that could match their skill sets.
Refuting the idea that robots will take jobs is a straw dog refutation. It is not simply robots but the convergence of developments in deep learning, quantum computing, algorithm mathematics, haptics, machine optics and robotics combined with the ever increasing divergence between the quality of education for the haves and have nots that portends an employment catastrophe.
1
yeah yeah yeah, everything krugman says here is good, correct thinking. However the actual problem is endemic: it is our love affair with growth economics. It seems to derive from the age of discovery, when there was always a new part of the world to find, and acquire. Always a lot of new resources to commandeer & exploit. For a few hundred years, an entire world to entice into consumption of 'stuff'. And of course now, when discovery is all but finished, its new tech. But the icing on this decayed cake is .. the new debt economy. The debt economy works via expansion enabled by borrowing. It's managed by allowing the value of money to decrease -- ie inflation.
The problem of course is that it's all self-defeating, absolutely not sustainable. What a debt society is actually doing is mortgaging itself, hoping no one is looking. Like the guy hiding behind the thin light pole.
3
I think that Krugman understates the risks of automation, robotics and AI. A group tried to revisit the definition of productivity a few years ago because of the fragmentation of the manufacturing among subcontractors, and they ended the meeting unable to satisfy themselves about how to accurately measure productivity. The questions that those who say not to worry include name 4-5 new professions that will be created as a result of automation, what IQ will be necessary to do the work, what education level will be necessary, how will these jobs pay relative to the ones eliminated, how long will it be before automation eliminates these new jobs, how many fewer people will be required to check the work of automation, e.g. radiologist and what if economic growth stagnates. This train is roaring down the tracks already and picking up speed. The question is when does crash into those who think they have a career ahead of them.
2
Dr. Krugman is absolutely right. The argument that automation will wipe out all the jobs have been around forever. I can remember arguing about this 30 and 40 years ago.
Yes, jobs are lost, but new ones pop up.
Thirty years ago, no one would have predicted all the information technology jobs we have today, which includes everything from white collar data analysts to the blue collar guys running the internet cables through buildings?
So where will the next jobs come from? I have no idea; if I did, I'd be investing in it like crazy.
When considering the workforce implications of rapidly advancing automation and artificial intelligence, Mr. Krugman and many of the commenters on this article seem to focus on so called blue collar jobs - manufacturing jobs, farm jobs, drivers, check-out clerks and the like. Mr. Krugman does briefly acknowledge that even white collar jobs will be impacted but I think this very much underestimates the scope of what AI will affect. AI will be capable of not only accelerating but improving lawyering, doctoring, authoring...even art making. Really, it’s difficult to identify any activity involving mental direction that will be immune. Not saying this is either good or bad but it is certainly disruptive. And it will further drive us to explore what it really means to be human. Hopefully, that knowledge will help us figure out how to live productively and harmoniously in a world where there is more than one form of intelligence - and wherein those intelligences may have different strengths (and weaknesses).
3
Can we at least investigate spreading the wealth accrued by this automation implemented? Currently our society is not structured to deal with this runaway growth in economic inequality. Because $ plays such a dominant role in determining a person’s value and self determination in our society, our benefits from equal opportunity and democratic gov. is rapidly being lost to this corrosive economic inequality. We are losing our freedom and our beautiful, thriving democracy. We need to address this dangerous problem yesterday!
3
Krugman is correct. Much of the developed world is facing a skilled worker shortage, including the US, Canada, Germany, and Japan. (You can Google it)
In my home state of Minnesota, the skilled worker shortage is looming as the largest threat to the state's continued economic growth. Vermont has instituted an incentive program to attract skilled workers. It pays for their relocation costs.
The "robot rabbit hole" is an apt phrase.
The problem is that people have a hard time fathoming that the telegraph, the railroad and the steam engine were more disruptive than the Internet. Electricity was more disruptive. True that AI and the mobile revolution is a big deal. But, folks, we've dealt with this before. We still have full employment.
2
It's the effect on people's world view that is the important thing behind UBi--not the actual cash. Many who support UBI see it as an imperfect and part-way solution to help society transition to a future that will very likely need to have different economic values based on a different economic reality.
A world in which much of the hard work is done by machines should be a good thing--leaving us time to pursue... well all the things we tell ourselves we might do if we had the time. Yesterday's NYT article about why rich people can't stop working so hard is a good look into the problems we face dealing with that future.
I know it all sounds very sci-fi--but the only alternative is to go backwards. It's also possible that this future hold many jobs we can't yet imagine. Still--better safe than sorry.
2
Krugman's third rail, or when Krugman goes crazy again. Still trying to rewrite his awfulness last cycle where he turned into the Democrats version of FOX news as did all the Times cognoscenti.
"The best answer, as I said, came from Sanders. No, I don’t support his proposed job guarantee, which probably isn’t workable. But he was right to say that there’s plenty of work to do in America, and right to call for large-scale public investment, which even mainstream economists have been advocating as a response to persistent economic weakness."
Had we the needed infra structure plan employment would be in a vacuum for up to twenty years.
Looking back, the Times played a leading roll in the 2016 disaster and Krugman played a definitive roll.
At the cost of putting Yang down Krug avoids the fact that the employment situation twenty year out look is NO JOBS!
Put that in your pipe (Paul) and smoke it.
3
@The Iconoclast Not true, we need people to grow the pot being smoked in Oregon.
Watch out Dr. Krugman, there's a baby in that bath water!
First, you assume a lot of flexibility on the part of displaced workers. Evidence from formerly industrialized northeastern states suggests that when the truck drivers are replaced by self-driving trucks, they won't happily retrain as baristas or masseurs, let alone software engineers on the coasts.
Second, there is substantial evidence that a universal basic income works to provide people security and flexibility in their career choices (see Bregmans' Utopia for Realists).
We don't have to be technological determinists to believe that our economy will not help workers to transition to new jobs, and a UBI is a great step towards fixing that.
5
The greatest disparity in wealth was not a product of the industrial age, the commercial revolution, the Third Wave, or any development since the time of the Enlightenment. It was the Age of Feudalism. Drug Gangs running a country is a throw back to Feudalism, surely a drag on the Mexican economy. Pseudo-Religious thugs running a country is a retreat to Feudalism, and ISIS is not known for gains in productivity or well-being. Accumulating economic assets into an inherited ruling class is a recoup of Feudalism, and hence the demise of unions and the American middle class, and the great and overwhelming cloud of economic insecurity and unending debt.
The tit for tat over capitalism and socialism is way off the mark. The enemy is Feudalism of the 21st Century.
4
My local Walmart has recently converted to mostly self checkout stations and I have read where they are laying off janitors and replacing them with robots. And autonomous vehicles are not pseudo-science. They are being tested to replace truck drivers. That is a lot of jobs disappearing.
1
Actually, Krugman’s right. AI/Industry 4.0 is far from what all the paranoia suggests. And it is best for repetitious/monotonous work. It still requires intelligent people who can (re)program, monitor, fix, maintain it.
Very simple analogy - think about all the old appliances in a home. I love my dishwasher but it still needs to be loaded, turned on when it’s full, and unloaded. When I do it, dishes come out clean, because I’m skilled at knowing how to fit things so that they all are exposed to the water jets and can drain properly. When anyone else in the house does it, not as much fits in, food doesn’t come off, dirty water puddles in bowls & cups, things get baked on again in the dry cycle, glasses are chipped, and utensils are still gunky. I need to make sure the drain isn’t clogged, the soap and rinse aid are filled, and it’s all maintained. BUT I have a dishwasher to save me time so I can do other things (more productive). I suppose I could choose to wash dishes by hand, but it’s more wasteful of time, energy, and water!
Full disclosure - I live in Silicon Valley, I’m an electrical engineer, and I spend a LOT of time working on Industry 4.0 and factory automation.
3
Normally, I agree with Dr. Krugman, but on this I disagree.
The problem with automation isn't that it is taking away jobs, but that we still treat jobs as a requirement for living, which is a feudalist/capitalist mindset. Automation removes the necessity for *productive* jobs, and in it's wake is a lot of non-productive work.
In the twitter comment he references, he notes that output per person is slowing in growth(not declining), but this is an easy mistake to make. Similarly, you could ask why we can't have a .5% drop in unemployment month over month forever, or why the marathon record used to drop by several minutes, but now it only drops by seconds. (Hint: It's not because top tier athletes are getting slower) There's a point where trending acceleration is minimized because it gets near to top end.
We are nearer and nearer to those margins, and more importantly, the available labor pool is expanding faster and faster. If we refuse to acknowledge that there is more available labor than there is work to be performed and don't adjust our policies, we will find ourselves in an even greater crisis of income inequality than we already have.
Piketty discussed this at length in Capital in the 21st Century, and as he warned, there needs to be a serious correction. We get closer and closer to that tipping point with no serious discussion about a correction. Don't stifle that talk, Dr. Krugman.
5
The relentless attention to the bottom line—with no weight given to human or community costs and benefits—has been used to justify the present absurd division of resources.
Warren’s tax on the rich would merely be a fair return on our collective investment in public infrastructure, education, and civic governance. Republican tax cuts to the “job creators” have yet again been gleefully sequestered instead of used as promised. Yang’s $1000 a month would at least go directly into the economic stream, lubricating and nourishing, the effect amplified by circulation.
I don’t hold much hope for the one percent recognizing the validity of the concept, let alone being insightful enough to reckon their recompense in terms of survival of the planet, but what a shame. I guess if you’re the blind Emperor of an ash heap, you can still tell yourself you’re on top.
3
I don't get it, how can we have inadequate private spending when corporate debt is sky high?
1
I think Mr. Krugman is oversimplifying the coming wave of AI based automation. It isn't robots doing mechanical jobs, replacing a technician. It is AI based reading systems that will displace pretty much all but a few attorneys and CPA's. The next round of 'automation' as Krugman refers to it, is coming after the white collar jobs with a vengeance. Apparently Paul isn't well informed enough nor technically educated enough to see the inevitable.
4
Sorry Professor ... I am usually in agreement with your positions, but I think your premises here are severely outdated.
A guaranteed minimum income is not "...an excuse for not supporting policies that would address the real causes of weak growth and soaring inequality..." Instead it is THE policy.
Yang, though not very articulate, is right. There is already a huge tech/productivity dividend being paid every year that is a result of decades of work by all Americans. Currently, however, that dividend is being paid almost entirely to the wealthiest.
Having that dividend paid to the people who earned it, IE The People, fixes many problems in one swoop.
First, it eliminates many of the social problems caused by the extensive poverty in this country.
Next, it makes us even more productive and efficient as people who really don't have much to contribute aren't forced to work. Imagine going to CVS, Home Depot, etc, with workers who actually wanted to be there and really were intent on providing the nest customer service?
Further, people would not have to choose between work and assistance. In the present system, if somebody manages to find a part time job, they often lose their "welfare" benefits.
Also, if everybody is receiving a base monthly income, we can do away with almost all of the present means-tested safety net ("welfare") programs. Not only would this save billions in administrative costs, but it would also end the paternalistic nature of these programs.
3
@Celeste
Yang not articulate? He may not ramble on like the rest (I'm guessing he's just naturally introverted, wouldn't that be refreshing after Trump?) and he's not the greatest debater (probably gets nervous and it's hard to cram in much during 30 second sound bites, much better if you watch his one-on-one interviews online), but he zeroes in on problems and solutions. Notice how pundits and other candidates try and shoot down his ideas, yet he has facts and reason to back himself up. I admire his clarity of thought compared to many of the others.
3
I encourage Krugman and others to read Yang's book and visit his website. His ideas go much deeper than just giving people money because of automation.
Many Americans struggle to pay their bills, dislike their jobs, and have no savings. UBI helps address all of these issues. It also gives value and dignity to unpaid work (caregivers, volunteers, etc.) and other life pursuits.
Many Americans (and Trump voters) don't trust government to spend money no matter what, even if it would help them or their communities (see recent NYT article, "In the Land of Self-Defeat"). So when Warren has a plan for everything, watch out. UBI, conversely, puts money, power, and decision-making in individuals' hands.
International aid organizations like "Give Directly" realize the value of giving power to individuals to make money decisions, not to government bureaucracy. UBI is simple and easy to understand and implement, unlike most government programs and taxes.
Yang argues that we measure the wrong things. A company sells weapons to a rogue state, GDP goes up. I stay home to take care of a disabled relative, no effect on GDP. An oil refinery has record profits, great for GDP (not so great for the environment). Someone wants to start a non-profit for at-risk youth? Not on GDP radar. UBI changes how we look at "work" and what we value.
UBI (along with universal health care) could transform our society in unique and profound ways, and warrants serious consideration.
7
Paul agreed with Bernie “he was right to say that there’s plenty of work to do in America, and right to call for large-scale public investment, which even mainstream economists have been advocating as a response to persistent economic weakness.”
What Paul has not pointed out clearly is that these jobs have to be in the public sector because corporate America sees no merit in the public good that helps everybody. Corporate America lobbies government only for competitive advantage that benefits particular corporate players.
Better healthcare, better education, better roads, better environment, ..., they help everyone and payoffs are long term. Not interesting to corporate America.
5
Capitalists are not heroic "job creators" who run around sprinkling employment opportunities about the countryside like Johnny Appleseed.
Anyone who has run a business with employees knows that one of the ways to be as profitable as possible is to employ as few people as possible and pay them as little as possible, while factoring in that you want happy employees who do a good job.
And that's not a crime. The point of a business is to make money.
But it doesn't mean that every employer's image should be cast in bronze and mounted in the town square to be genuflected at.
Employers are gonna do what employers have to do. Robots are not the problem. A planet drive by zero-sum capitalism is.
1
Re: "an inadequate solution to an imaginary problem" That is a nice quip, but remember that the issue is not just about the *quantity* of work and not only about income, it is about the *quality* of the work
Though different people have different notions about what they want from work besides pay, it is generally true that people want meaningful "work". For some that means a simple, low-stress job working reasonable hours so that one's fulfillment - one's meaningful "work" - is centered on non-wage activities (family, spiritual, artistic, intellectual, hedonistic, etc.). Others seek meaning (and status and sense of place) within their paid occupation.
But nobody likes a terrible, tedious job. And this is an important thing to look for: Is a form (or set of) automation liberating us or enslaving us? Two technologies may print the same productivity number but have very different "unmeasured" impacts.
Yang's proposal completely misses the mark and says volumes about the massive blind spot of the supposed tech elite.
Of course the politicians have no idea about the automation because they have never worked in an industry where automation replaces humans.
Automation has been around since the dawn of human history, every time a tool was invented and it replaced a part of human work. Henry Ford started the factory which eliminated many labour jobs from the factory but it also created many jobs elsewhere because people can move and transport goods.
So the end result is that jobs are simply moved from one industry or one place to another place. Obviously you hear a lots of complaints from the people who lost their jobs but no voices from those who gain(they are busy working instead).
Politicians are trying to gain votes and they are doing everything to show their compassion, and all that talk about automation is just to win votes and get elected. Once elected, it's business as usual, and those people lost their jobs are promptly forgotten.
Look no further than Donald, the 5th grader...
There is a big difference between the previous technological innovations and this one.
Previously there was always "something else to do". But intelligent machines (strong AI) will be able to do everything. This is not a problem for the next few years, but definitely for sometime this century.
One needs to think about the solutions to this.
2
Governor Steve Bullock has, from day 1 of his campaign, been focusing on exactly what this article calls for: pragmatic, achievable solutions to the real challenges of our time. He’s polling @ 4% in Iowa now, ahead of 7 candidates on Tuesdays stage, and his fundraising ranks 7th. He came out of the one debate he appeared in as a “winner” and was very effective in highlighting his progressive record of achievement as a Dem Governor in a red state. He is the only Governor remaining in the race (no senator has ever unseated an incumbent, but two governors of small states have in our lifetimes), and the only candidate to win statewide in a state also won by President Trump. Interestingly, every day I see these candidates actually following Governor Bullock’s lead as he progresses in his campaign with solid policy proposals and fair, workable solutions. I look forward to watching Governor Bullock’s continued emergence and I hope The Times gives him the coverage he deserves as one of the Democratic Party’s great leaders, and its most promising candidate for President.
"By 2005 or so, it will become clear that the Internet's impact on the economy has been no greater than the fax machine's." - Paul Krugman 1998
I respect you greatly for your achievements in Economics - but Dr.Krugman, you are so far off the mark when it comes to technology. This isn't just future problems, this is problems that have been going on now. You do not need FULL automation to have catastrophic effects to Americans and communities.
6
Andrew,
The early internet, with it's pre-browser interface was used by a relatively small group of highly educated techies. It was the windows-based browser that leveraged the potential power of the internet for millions. It is the difference between windows and DOS. With DOS a user had to know fairly arcane commands to use their computer, with windows graphic interface it was point and click.
1
@joe parrott In 1998 I was playing Starcraft and Age of Empires on my Macintosh, talking to strangers not just through games but on AOL Instant Messenger. And that was just me as a kid!
1
The author of Sapiens and Homo Deus argues that automation has cost us more jobs than globalization, so apparently what the data says depends on who's looking. Krugman, Sanders and Warren are probably right. Globalization at best has traded good paying jobs in manufacturing for low paying jobs in services in exchange for slightly lower retail costs. That's great for people who still have jobs but not so great for those who don't. Also, anyone who drives across this country will encounter small towns that are dead or dying because their job base moved to China or Mexico. The effects have been catastrophic.
All this makes me think the foremost objective in an American economy should be to employ Americans with good jobs and decent wages. If it takes a mixture of protectionism and federal spending to sustain a middle-income working class, so be it. The investor class will be fine either way.
1
Big Krugman fan here. His articles are clear and thoughtful. The mistakes I have observed him make are few and clustered. His famous guess that the Internet would not be a big factor in economy was very off, and of the type of error he may be prone to make. The poor guess that the market would react badly to a Trump was another.
In general, economics relies a great deal on historical observations. The reliance on correlation as theory makes it difficult for economists (and all of us) to account for substantial technology or political shifts.
The hype of the computer science crowd is real, too. Anyone remember the 5th Generation, and Y2K? Coursera would end teaching as we know it?
Any of the Democrats will be better stewards of the government than Trump should such a shift be on the way. Even Warren and Sanders, who side with Krugman over Yang, will respond thoughtfully should the transformation be real.
There is so much left undone across the globe. Think about the uneducated hungry souls across much of Africa, parts of Asia and the ME, as well as parts of Central & South America. Here too, our fellow Americans in certain parts of the U.S. are hungry and poorly educated.
The climate crisis, wide spread corrupt governments, currently including the U.S., poor food and water quality. I could continue but the point is made.
The next generation will be productive with plenty of work, yet different work that does not register in current economic measures. What will change is the current economic model. To that point, Yang is blindly heading in the right direction, and Sanders & Warren are even further down the road towards figuring out this necessary new economic roadmap.
2
Occasionally my wife and I have to go to Syracuse NY to take care of her ailing uncle. On the way to the oncology center, we drive past GM circle, where once resided a large GM plant that employed her uncle, on to Carrier Circle once home to a giant carrier plant that employed her late farther, no bulldozed flat, green grass hills indicating where the buildings where flattened , We then move on to New Venture Gear drive, once a large company making transmission of all of Detroit’s big three, now gone, moving past Chrysler drive, gone too
My wifes relative lost their jobs to the companies moving to Mexico, after NAFTA was signed -
Not robots.
It is time for the Democrats to address the real cause
3
Free trade is one of the major reasons many good paying jobs went away. Automation is a large part of the reason that, even with the most protectionist trade policy possible, they are not coming back, at least in large numbers. But these are not the only reasons. So it seems to me that possible remedies would need to hit the problem from several different angles at once. For starters, raise the minimum wage, increase the power of organized labor, and crack down hard on employers who use illegal labor or other shady means to depress wages. This should help push those low paying jobs that can't be easily off-shored or automated closer to a living wage. Raise taxes on those, myself included, who can afford to pay a bit more. And the wealthy should be taxed at especially high rates, in order to catch them up for the relatively cheap ride they've been enjoying over the past 40 years. Think of it as back dues. Supplement this with internet fees, as Yang suggests, as well as transaction taxes on stock trades. And eliminate the cap on payroll taxes, as well as making them applicable to all forms of income, not just wages earned through actual work. This can fund much needed infrastructure projects, with participating contractors forced to pay a living wage, as well as investments in education, and some form of universal health care. I think Medicare for all is probably the best, but I suppose we shouldn't let the perfect be the enemy of the good.
2
The rapid wage growth and expansion of the middle class in the 1950's was primarily brought about by the GI Bill. If the American Legion still had political clout, they could lobby for what is needed now, a "Tech" GI Bill.
2
Occasionally my wife and I have to go to Syracuse NY to take care of her ailing uncle. On the way to the oncology center, we drive past GM circle, where once resided a large GM plant that employed her uncle, on to Carrier Circle once home to a giant carrier plant that employed her late farther, no bulldozed flat, green grass hills indicating where the buildings where flattened , We then move on to New Venture Gear drive, once a large company making transmission of all of Detroit’s big three, now gone, moving past Chrysler drive, gone too
My wifes relative lost their jobs to the companies moving to Mexico, after NAFTA was signed
Not robots.
It is time for the Democrats to address the real cause
There is a small but vital road bridge near here. It is simple ferro-concrete, nothing fancy, but we wouldn't be driving over that small stream without it.
We've been driving over it since my grandfather was raising young children. It has a clear sign on it, that it was built by the CCC during the Depression. That is the Civilian Conservation Corp, that did needed public work to get the country back to work during the Depression, to "prime the pump."
That bridge won't last forever. We'll need to do it again.
It has certainly paid us back for the effort. It was a good investment, besides helping to end the Depression.
That FDR knew what he was doing.
3
During a recent trip from Wuhan down through the Yangtze Valley Economic Development Zone, into Anhui and Zhejiang, I was able to see the large number of factories and unbelievably vast industrial infrastructure that is being brought online for Xi’s attempted push to dominate the manufacturing sector. We stayed in a converted silk weaving manufactory. The owners were a family that were, at one time, medium sized players in the silk business. They converted their facilities to tourism due to the influx and expansion of larger players in the silk business who populated their factories with robots instead of humans. Villagers now go into the cities as migrant laborers.
Another trip into Fujian, talk was of how all the factories have closed, with large numbers relocating to Vietnam and other SE Asia countries, and the consolidation of manufacturing activities to the central China cities of Wuhan and Chengdu, with the new shops adopting robotic technologies instead of workers. I have been in the vast industrial parks in Optics Valley, Wuchang, with vast manufacturing facilities and teeny tiny little parking lots because....hmmmm...Why would there be indescribably huge manufacturing facilities and infrastructure but only teeny tiny little parking lots? Is it because of the excellent public transit that brings workers to the site or....something else?
China is the lab for seeing where Industrial Capitalism is headed. Paul might want to take a look before he writes.
4
The key point in the article is that automation has been around in its current form since the industrial revolution.
Societies and cultures need to adapt to automation.
But Yang has gone off the rails with the monthly stipend philosophy. It is much too low and not means tested.
Trade is the main reason for job loss to countries that can manufacture products and provide services at a lower price point.
1
The business I've known longest is lawyers. They've been very exposed to automation and change. Legal research, writing, and methods to check knowledge of the law is entirely different from when I started. It is so different it is unrecognizable.
Are there fewer lawyers because of that? Hardly.
Exposure to automation increases productivity. Increased productivity is not the same thing as increased unemployment. Certainly no for lawyers.
Not for things either. Cars are bigger and better. They are unimaginable compared to my beloved 1968 Mustang. There are not fewer cars. More people have more cars. Computers in them, computers designing them, computers running assembly, have just made better cars and more of them.
What about that assembly of cars? Don't confuse jobs no longer done with jobs sent overseas for someone else to do. Somebody still makes those cars. They have a higher labor input, not a lower input.
That higher labor input is higher tech too. Instead of lifting a wrench to a frame moving down the assembly line belt, they run computers that run robots that flash weld on that line.
My father knew a guy whose right arm was monstrously disproportionate to his left, from raising the same heavy lifting lever so many times every hour of every day. Today, his equivalent sits in a hard hat behind a protective clear screen and runs the robot.
We are doing so much more, better cars, better everything, that the labor is needed, even with all the available automation.
1
i would enjoy reading the lunch conversation between david brooks and mr. krugman on the topic of liz warren's "realism."
a careful reader will find that mr. krugman changes the focus and relevance of examples to his argument many times. this does not signal evasive reasoning but reasoning about a badly, vaguely formed question.
the reason mr. krugman veers from trade to productivity to wealth inequality to politics is that he is, like a solitary blind man, feeling all parts of the proverbial elephant. the nature of the beast remains obscure.
i believe that robotic cars and robotic surgeries are far in the future. but i know that our network of networks, our power grids, our financial markets, our logistical systems of procurement and distribution are already under AI or "robotic" control. boeing is on the ropes because its robotic airliners seem to like flying into the ground. concerns about "automation" are valid.
human is a biological fact, and the basic issue is that a biological fact is evolving greater and greater dependence on a vast, digitally guided and carbon powered machine. yang and others are right to be cautious about that prospect; warren and others are right to call out its wealth dynamics and consequences.
but the focus should remain squarely fixed on the machine itself: built for profit, powered by carbon, lavishing the herd animal with the illusion of individual importance.
1
When lamenting the loss of jobs to innovation, those lamenting forget all about the new jobs that are created. Remember, there were no PC's in the fifties, there were no cell phones in the seventies, nor were their gaming platforms in the sixties. All those phones, computers and gaming systems have to be designed, constructed, marketed, sold and shipped. Televisions continue to get better, which is attributed to the components of those televisions getting better. New medicines and analytical devices are also being invented. We need researchers for that, and when they are successful products, we need to produce them too. Drilling for oil in the US was a dead-end just a decade ago, but fracking turned it into a boom. Coal is dying too, but natural gas, solar and wind energy is booming. It's all just change, folks, so get used to it. The human mind never idles - at least not collectively. If we run out of new things to do here, we'll just have to go to another planet and we're already working on that.
Mr. Krugman’s effort to equate past industrial revolutions to what is called the 4th Industrial revolution is a mistake. Previous industrial revolutions were primarily focused on the efficiency of human work. This revolution is focused on the efficiency of human intelligence. There is a good reason they call it artificial intelligence (AI) and not artificial work.
For the first time in history, machines can learn, comprehend, analysis, create, and evaluate faster than their human creators. To equate this development to working on a farm 100 years ago is simple nonsense. AI is not the end of the world, but it does require societies to prepare for the inevitable changes that will occur.
We must be honest with ourselves and accept AI is coming and many things will change, but we must also accept the responsibility to shape that future. Mr. Krugman’s piece only serves the few for-profit companies who want to shape the future without acknowledging detrimental consequences. If we accept Mr. Krugman’s premise, we all will be serving these same companies. Unfortunately, we are well on our way.
1
I've been a computer programmer for 40years.
I've worked in industrial automation, record keeping, and environmental monitoring. I'm sure positions have been eliminated by the work I do. Mainly this impacted clerks, material handlers, and such. My work has also created jobs for supply chain managers, order processors, logistics workers. These jobs couldn't have been done fast enough with a room full of people coordinating deliveries, orders, etc..
I have been replaced twice in my career by cheaper programmers from India. Retrain, Retool and move on. No,it wasn't easy, yes it involved hardship, but that's how it goes.
With increased automation, AI, robots the job creation/destruction cycle will continue unabated.
Life isn't static, it hasn't been so for the last 100 years. There's really only been 2, maybe 3 generations throughout human history that could count on a lifetime job (other than being a farmer). That was the exception, not the rule.
Fear is always an easy sell.
2
Absolutely agree...democrats are getting to wrapped up in side-stream eddy's. Concentrate on the river of health care, jobs, and wealth inequality...
1
Interesting column, to a degree. Let's face it, Yang might make a great dinner guest, but he isn't going to be the next president. Furthermore, most of what he brings to the debate is but a sideshow. Consider this, what many consider to be the most imminent threat to survival of the planet didn't get but a whisper. I believe that might also have some economic implications.
2
Krugman gave a really interesting perspective here, but I think he neglects the key synthesis that pulls everything together here; productivity may be growing slower than it grew historically, but the spread between productivity and wage growth is at unprecedented levels.
Automation is not to blame for this unprecedented spread. Oligopolistic labor markets are to blame, and that has been Warren's position for ages.
504
@AB I agree (for the most part). What Krugman and Yang both miss is that if you actually give workers/stakeholders a say in how our economy is run, the whole automation problem disappears.
Automation is NEVER a force of nature—technological improvement could very well be used to EMPOWER workers. But technology is ultimately financed, developed, and owned by the same class of people (billionaires/multinational corps) that own the economy writ large. Why invest in workers when you can replace them with robots and siphon off their old wages into executive pay, dividends, and stock buybacks?
Automation, wage stagnation, offshoring of production...all of these problems will go away very quickly if we just attack inequality at its source by democratizing industry.
31
@AB
I think your comment is right on the mark. Non-technical people have this mental vision (supported by the media) of robot welders making cars that used to be welded by people so, the logic goes, all welders must be out of work.
Robots aren't cyborgs who compete with humans; robots need people to design them, install them, maintain them, and program them. The advantage of robots is that they increase productivity but, as AB says, it what has happened to the increased profits that come from increased productivity; it's all gone into the oligarchs' pockets and increasing wealth imbalance is the most obvious symptom of this problem.
We shouldn't just be talking about minimum wages, we should also tackle maximum wages. The Economic Policy Institute epo.org has this to say...
CEO compensation has grown 940% since 1978
Typical worker compensation has risen only 12% during that time.
I rest my case.
36
@AB....Right.
We constantly hear about automation. But our media is neglecting the truth.
Regardless of any automation--the big fact staring at us is that our elected lawmakers ok’d the planned, deliberate offshoring of millions of our jobs. Factores closed, towns were ruined.
This was designed to up corporate profits and thus mega donations to our parties. This has ruined our politics —big business sets the norms and limits, calls the shots, defining what's too 'left' --for its gain and our loss.
We The People can't compete.
Where is Krugman and the media on this?
The media looks good bashing/exposing Trump, easily the worst president in our history. Trump has found our economy and our politics to be easy pickings, ripe for exploitation.
8
Fools rush in, but....
Isn't "productivity" a measure, not of the amount of stuff a worker can produce in an hour, but rather the value of that stuff? And isn't the value (the price, that is) likely to fall, perhaps by a lot, if the cost of production falls?
So a worker making twice as many widgets per hour, where the price of the widgets gets cut in half (I'm ignoring TPF for purposes of this rant), has achieved exactly zero gains in "productivity"?
And where the worker produces the same number of widgets, but way better quality ones, for the same price, also achieves no gains in "productivity"?
And when a company decides to provide everyone in the world with something that used to be scarce and thus valuable (maps, Candy Crush, GPS systems, etc.), it actually reduces "productivity" because no one is paying for those things any more?
I'm sure I must be wrong about this, because Sweden, but where is my error?
Democrats will never own up to it but the Clinton NAFTA concoction was one the largest job killer phenoms in recent history. The TPP would have been the 2nd largest job killer in America had it been enacted. Trade pacts dressed up as free trade that send jobs to low wage countries are JOB EXPORTERS a.k.a. job killers. Trump is giving tariffs a bad name because well placed judicious and strategic tariffs could have saved millions of jobs over the last few decades.
2
Burnett isn’t exactly the sharpest knife in the cutlery drawer, nor are almost all the rest of the moderators of the debates. She’s telegenic, as are most of them. That’s one of the biggest problems with the organization and format of them. They are television personalities first and journalists second; they try to command the spotlight even though—or perhaps because—they’re fundamentally irrelevant to the very purpose of the debates. The public is the loser.
1
My fifth grade teacher would constantly admonish us: "You can be replaced by a button." After sixty years, it is nice to have this menaching childhood idea confronted by a grownup. Thanks.
The election of Donald Trump is not an "imaginary problem," just because Clinton supporters were too closed-minded to see it coming.
1
Hal 9000 was built at the University of Illinois Champaign Urbana.
Hal was a fictional entity from the film "2001, A Space Odyssey' with a real world warning about the threat posed by human avarice and hubris regarding science and technology.
Our science and technology frequently runs ahead of our educational, legal, moral, political and political ability to properly identify human individual and societal costs and benefits before it is too late.
See 'Frankenstein; or The Modern Prometheus' Mary Shelley; 'The Strange Case of Dr. Jekyll and Mr. Hyde' Robert Louis Stevenson; 'The Island of Dr. Moreau' H.G. Wells
1
@Blackmamba
3rd paragraph ...replace 2nd political with 'socioeconomic..'
So, once again I agree with Paul's sensible arguments and disagree with his backhanded dismissal of Bernie Sanders. In Paul's world, the world of neo-liberalism, Bernie is a wild-eyed socialist. In my world Bernie is the champion of completing the transition to the New Deal begun 85 years ago by FDR-- universal health care, a provision provided by every Western European country, stronger social security, strengthening the position of workers in of all places the work-place, and so on. Yes, Bernie got it right, the problem is not that there is a lot of urgent work to do, it is that the economic system we've constructed does not get the necessary work done. Why? As Bernie points out profiteers have taken over the government and use it to milk the 90% for every more obscene concentrations of wealth. Society's productive resources are not focused on building sustainable industry, protecting our precious environment, caring for people's mental and physical health. They are squandered on finance and an increasingly monopolized money sucking endeavors rather than creating real wealth. I, like Paul and a lot of others, am a fool. I backed Clinton in 92 and 2016 for the wrong reasons, a feeling that incrementalism was the only hope. In fact it just convinced voters that we were going to get more of the same long, demoralizing slide into plutocracy and environmental collapse. Paul still can't see the obvious fact that Bernie offers the most realistic way forward.
3
The awful part of these "debates," as suggested, is that candidates have to accept the PREMISE of a question, no matter how ridiculous.
1
So now robots are the true cause for the demise of the working class.
It seems like just yesterday that the true cause was lazy, overpaid, under-educated union workers.
Both explanations have much in common. Each is based on flawed, if not bogus, theory supported by cherry-picked, if not fabricated, data presented to the public as divine truth from the gods.
Let's call the new explanation what it truly is - a lie packaged for delivery in spiffy sound bites to confuse and deceive the losers so that the pillaging continues and the liars have time to formulate another explanation to justify the inequalities.
And somehow the policies that favor capital over workers is never even mentioned.
2
I can envision robots becoming smarter than people and one day looking at people and seeing people as a problem that needs to be eliminated as those greedy people pollute and destroy the earth.
1
Automation of work is a threat? What about automated politics? And I don't mean primitive robotic phone calls from candidates or PACs.
Robots may never vote or run for office but the reality is that targeting of voters, big data-driven manipulation, social media bots, voter profiling, issue framing, message development, even candidate selection, are already dominated by algorithms.
What Chomsky called the Manufacture of Consent was a crude preview of technology that captures attention, pushes precise hot buttons, twists perceptions, savages facts, distorts context, inflames ignorance, costumes lies as truth and traitors as patriots.
The means of surreptitiously identifying, engaging, exciting and mobilizing a significant political base may seem consistent with democracy but in fact it subverts the role of citizens from engaged and rational political actors to a passive audience enthralled to impulse gratification and cheap spectacle.
It's the difference between Nature on PBS and the Gong Show on You Tube.
Politics isn't about persuasion or winning hearts and minds any more. In the automated version of democracy, the entire enterprise is mobilization of a like-minded "Us" (identified by past votes or markers like NRA or ACLU membership, google searches, website click counts, credit card use, Starbucks or Bass Pro Shops) against a big-data designed "Them".
Automated politics creates the illusion of democracy in a gilded cage of inequality and injustice.
3
Are you an Automation Denialist? Don't you "believe" in "science"? It's an eye opener how much we don't know about history and if we do we misinterpret it. People have been moving off farms and into villages and from village to village since the before the Ice age. We're living through the last phase of the industrial revolution which started over 250 years ago but the AI revolution is going to happen much faster. Some big brains say thirty years. The unintended consequences of the Industrial Revolution like Democracy, Communism, Nazism and two world wars were a pretty high prices to pay to say forget about it.
Look at it this way people were saying the middle class was disappearing 30 or 50 years ago. What a rabbit hole for the Democratic Party to get involved in, so they didn't, and today people are voting for Trump. But I guess if you're on your way to greater and greater wealth it won't matter.
"...Warren was surely wrong to suggest, however, that changing trade policy would do much to bring good jobs back. "
Uh?? Why? Krugman's vestigial 'Trade Good! Trade Good!' muscle memory is kicking in here.
Wasn't it American legislation that made possible for American CEOs to move most of manufacturing to China to leverage low labor cost, increase profits, pump up the stock and put more money into their pockets? Krugman is being breathtakingly disingenuous in ignoring where China sells all its products. Is it the American market or Timbuctoo? The Trump tariffs gave the Chinese economy pneumonia. The simple way to bring manufacturing back is to make it unprofitable for American CEOs to manufacture in China via tariffs and tax policy. We should also nationalize banking to make it a public utility. Of what good have money lenders been over the ages? Jesus Christ threw them out of the Temple, Bonaparte trashed them as a Nationless band of greedy Jackals. Wall Street has always been a viper in the bosom of this great nation and it is time to cut off its head.
1
"...and the debaters by and large — with the perhaps surprising exception of Bernie Sanders — gave pretty bad answers."
Only surprising to you, Paul. The rest of us are really glad Hillary is on the sidelines this time.
1
I am left to wonder how many times Sanders will have to give sound and sensible answers to economic questions before Mr. Krugman finally has a more positive opinion of him.
4
When I read headlines that Obama won Ohio, and we all need to remember how that was done, did everybody forget he lost Ohio primary in ‘08? How was that done?
Simple. Super bad timing for him. At the time, he was caught telling the Canadians through some seriously sloppy back-channel talks with Canadian contemporaries that he really didn’t care about NAFTA.
Why did Hillary lose in ‘16?
NAFTA.
What is NAFTA?
It is shorthand for very bad things, regardless of the jobs it encouraged and the markets it strengthened.
It is a scapegoat. There’s no nuance tolerated anymore. BTW, that is super bad for pro-NAFTA Biden. Anyhow...
Enter robots as another scapegoat or symbol b/c it’s another gut over brain issue. I think it’s okay to play this by how people feel and not risk lecturing them on why they are wrong.
In the end, Obama didn’t need Ohio when he got 1/2 the delegates for the win in two states where he was not even on the ballot, so, what do I know?
1
While Mr Yang seems like a nice guy, I’ve always thought he used automation, along with the $1,000. handout, just to grab attention in order to have something to say in his debate performances. As always, Paul, your facts and knowledge speak volume to talking points. I so hope the other Democratic candidates stay out of Mr. Yang’s rabbit hole, and deal directly with the myriad issues our brutalized country faces, made far worse by the stupidity, criminality and divisive hated of this repulsive administration.
2
Watching the Democratic primary debates unfold Bernie reminds me of Bibi Netanyahu. Both have contributed much to their country’s well being but their time has passed. Netanyahu needs to cut a deal with Gantz and move on and Sanders needs to do the same with Warren.
Nothing like cold logic and hard facts to debunk a myth. Very well done sir!
what a wonderful, sensible column!
Yes, it’s amazing that this pseudo-problem got a debate question and substantial airtime and climate change didn’t get either. I can promise you right now that climate change is going to cause way more chaos, and be way more expensive, destabilizing and damaging than any robot revolution.
Millions of our jobs have been off shored to low wage nations.
Quotes from transcript of PBS Frontline, written by NYT HEDRICK SMITH--- “Is Walmart Good for America?”: His book is titled “Who Lost the American Dream?”
"At the other end of the pipeline, I visited the port of Long Beach, CA, to see how Washington's promise of massive American-made exports to China was working out. The port's communications director is Yvonne Smith.
HEDRICK SMITH: What are they shipping in and what are we shipping back?
YVONNE SMITH: Well, we're bringing in consumer products. We're bringing in about $36 billion worth of machinery, toys, clothing, footwear.
HEDRICK SMITH: That's $36 billion right here in Long Beach?
YVONNE SMITH: About $36 billion comes through Long Beach from China alone– consumer products.
HEDRICK SMITH: And what are we shipping back?
YVONNE SMITH: We're shipping out about $3 billion worth of raw materials. We export cotton, bring in clothing. Export hides, bring in shoes. Export scrap metal, bring back machinery.
HEDRICK SMITH: So they're doing all the– we're like a third world country.
YVONNE SMITH: We're exporting waste paper, containers full of waste paper. We bring back cardboard boxes with products inside them.
HEDRICK SMITH: Add it all up and the U.S. had a record $120 billion trade deficit with China last year, and it's headed even higher this year."
What's the solution to above?
2
" ...Sanders...was right to say that there’s plenty of work to do in America, and right to call for large-scale public investment, which even mainstream economists have been advocating as a response to persistent economic weakness."
EXACTLY! Hurrah for Sanders and Krugman!!!!!!!!!!!!!!
The US will stop investing (capitalist) in the worthless money that those who have it have hoarded away in US Treasuries, domestic and overseas bank accounts, and likes of Mar a Lago real estate and be spending (socialist) on infrastructure including transportation, health care, and education among all the citizens in the coming recovery.
A progressive tax will pay for the government spending those worthless dollars hoarded by the super rich for the not so rich.
How much longer will Americans need to keep shoveling their earning up the income scale into the dark holes of the superrich.
Wake up; enough already!
1
This is the second article I've seen written that completely disregards the AP fact check -- I'll trust the integrity of one of the most influential publishers more than an opinion columnist, thank you.
Congratulations, Paul, on your first column which is both reasonable and correct. More like this please!
Krugman, commenting on 9/11:
"The terror attack could even do some economic good... The driving force behind the economic slowdown has been a plunge in business investment. Now, all of a sudden, we need some new office buildings.”
2
Yang and Gabbard are fake Democratic candidates. The media gives them a platform a million times bigger than their actual political footprints. They don't deserve our attention.
I’m confused by your criticism of the centrists. It’s obvious we’ve lost and are losing jobs to automation (as well as global trade). All you have to do is walk through a US factory or think about truck drivers & self driving cars to know that. Or you can look at any number of balances & respected studies outside the realm of the “media’s robot talk”. Here’s one, see pg 23 - 28.
https://www.brookings.edu/wp-content/uploads/2019/01/2019.01_BrookingsMetro_Automation-AI_Report_Muro-Maxim-Whiton-FINAL-version.pdf
You don’t propose a solution to income inequality (which I agree is a huge problem) except to blame ”private spending.” Do you mean consumers buying TV’s? I don’t see how that helps income inequality when most consumer products are manufactured by using lower-cost strategies like automation and low-cost foreign labor. The more Americans who can be brought out of underemployment (Walmart) with either training/higher paying jobs or some kind of UBI the more TV’s we’ll buy.... We need a multi-pronged approach to raise middle class income: education/job training, better safety net benefits, Canada-type job-focused taxation of industries benefiting from global trade, etc.
I usually enjoy reading your perspective but this was just a “whine” with no proposals for a fix.
I disagree with Mr. Krugman. Automation is a real threat to American jobs. Watch Tony Seba talk about technological disruption, and then revise your statement.
1
"Warren was surely wrong to suggest, however, that changing trade policy would do much to bring good jobs back."
I was wondering when and if he was going to answer Warren. On the UAW picket line a couple of weeks ago she sprung her line about moving jobs out of the country:
"Their only loyalty is to their own bottom line. And if they can save a nickel by moving a job to Mexico or to Asia or to anywhere else on this planet, they will do it.” (NYT, Sept. 22, 2019)
That "nickel" represents the fruits of globalism in all its glory - something Krugman has repeatedly promoted as the way forward.
Now comes Krugman saying "we" got it wrong regarding the impact of globalization on manufacturing employment. Sorry Krugman, YOU got it wrong. Now you want to advise the Democratic candidates who by definition have to pander to the labor mobs?
From Krugman's mea culpa in Bloomberg Opinion:
"To make partial excuses for those of us who failed to consider these issues 25 years ago, at the time we had no way to know that either the hyperglobalization that began in the 1990 or the trade-deficit surge a decade later were going to happen."
Are you kidding? The globalization writing was on the wall in the late 1980s. And trade deficits? Well they really don't matter that much because money flows back to the US in the form of real estate purchases, right?
1
Outlaw washing machines, that will fix things.
1)Missing here is the fact that the industrial revolution one hundred years ago caused massive unrest and riots. So to say that we got through it even better than before... We'll talk to those who des. I'm sure they'd have a different take. Money would likely save lives.
2) going from horse to the model t is not the same league as mechanical machines to intelligent ai. It's not even the same sport.. one still requires humans. The other doesn't. Do we really think Amazon is trying to eliminate it's workforce with ai because they're dumb? Or because it would save them billions.
3)Paul, Andrew invited you to tour a factory with him to show you the automation taking place. Have you taken him up on his offer??
P. Krugman 1998, “The growth of the Internet will slow drastically, as the flaw in ‘Metcalfe’s law' becomes apparent: most people have nothing to say to each other! By 2005, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s”
2
Dr Krugman,
Thank you for pointing this out.
Many educated liberals try to show off their smartness w/o realizing the effect of their analysis-paralysis and confusion over most voters.. people who pays less attention or have less attention will remember lost jobs.. nothing else..
This is what killed Hillary and may hurt Dem chances to try to provide a deep coherent explanation.. which ii simple American mind - call defending.. Which is a loser in today's America.
Automation consigns the poor to beggardom, the middle class to poverty, and the rich...to ever greater wealth.
1
In my MA town, everyone has boycotted and complained loudly about the job-killing self checkout machines.
They have disappeared.
1
"They should focus on the real issues, and not get sidetracked by the pseudo-issue of automation."
Really? Tell that to the hundreds of thousands of truck drivers who will lose their jobs when autonomous vehicles become ubiquitous.
Tell it to the thousands of people who haven't been hired as cashiers at grocery stores and other retailers because companies do everything they can to force shoppers to use self-checkout lanes. How do they force shoppers to use self-checkouts? By refusing to hire cashiers.
I would have expected more from someone who received a Nobel Prize than a revised version of the tired "buggy whip" allusion, but I guess laziness is an outgrowth of sitting in an ivory tower secure in the knowledge that a robot is not going to take over the task of cranking out tripe like this column--at least not yet.
1
Dr. Krugman, I often agree with your perspectives, but on this subject, I believe that you are flat wrong.
I am a graduate student in the field of machine learning. Something you are not seeing from your distance to this field is the rapacious hunger to automate everything, because the winners become overnight millionaires.
Farming drones, self-driving trucks (most common job in most states), robotic factories, the list goes on. Plus, it's not just cashiers - the big bucks go to automating college-level jobs. And don't even get me started on synthetic media, which is already approaching photorealism; AI is going to hit the entertainment and media industries like a sledgehammer.
Furthermore, we don't walk towards AI - we RUN. Every major government on the planet is working on autonomous military hardware, and the best killer robots will be as advantageous in warfare as Nazi tanks were against Polish cavalry. Take away the guns, and you have agile, rugged robots with impressive object recognition and decision making capabilities with more civilian uses than I can think of.
There is no slowing down this process, not for anybody. Many of the most significant advancements in AI were made in the last SEVEN YEARS since the ImageNet breakthrough. Do you really think the next seven years will look the same as the past?
Ignore the thinking machine at your own risk. I jumped into this field to protect myself from the future. NYT columnists may be safe, but reader, are you?
2
Please dig into how regressively (VAT) funded UBI which can only be received if you agree to give up welfare benefits is definitely not the answer.
First of all, Paul Krugman, why are you surprised that Bernie Sanders gave a good answer?!
1
I agree. Automation is not the problem. The problem is who has the power to reap the rewards from automation, from International trade, from technological advances? The wealthy elite have been rewriting the 'rules of the road' for decades so that they always win and they always get more of everything; income, wealth, education, but most importantly, power.
https://whorulesamerica.ucsc.edu
2
I really like Paul and agree with most of what he writes. Unfortunately, he has been consistently wrong about the automation argument for years. The most recent Democratic debate just gave him another opportunity to trot out his same old talking points about worker productivity data not reflecting what he expects it to show if automation was happening. Therefore, he’s right and we’re all idiots. Sadly, he’s forgetting (or ignoring) the Productivity Paradox (“Solow Computer Paradox”) and Robert Solow’s famous quote from 1987 that “You can see the computer age everywhere but in the productivity statistics.”
Maybe he’s right and the rest of us are wrong. But maybe he should be a bit more humble and check his assumptions about where and how automation will show up in the data, economic or otherwise.
1
I think the biggest problem is that we're not being good to each other.
If we're good to each other, then the robots don't pose a risk. But if we're bad to each other, which we seem to be at the moment, it poses incalculable risk to everyone.
We don't have a society that can deal with this. Bernie Sanders thinks socialism can solve it, but I disagree.
We're treating each other worse and worse. We're sure that the neighbor is just quite as good as us. And we're wrong! The neighbor is a good person in 99 of 100 cases.
Our leader is nuts. Trump is a looney toon character. I've often posted the classic toon of the rednecks, it's really funny.
He's not acting funny..
And I'm prould to say my gal is winning! Liz Warren, bless your courage!
I love you, Liz.
Paul, I know you have written many books that not very many people read.
Only people like you, who are so far removed from the realities that everyday normal people face, claim that automation is fantasy. Remember the time NYT predicted Hillary's chance of winning against Trump was 98%? Yeah, what happened there? Trump won many of the states in part because the voters in those states lost jobs due to automation.
Automation will only accelerate and more jobs will be lost because of it. You should write a book about how wrong you were in 10 years. That one will probably sell more than your previous ones.
Imaginary problem except all the data says it’s a real problem. That’s why this is an opinion piece, it’s fiction.
1
Nicely written and I agree.
In a nutshell here’s the answer to our problems:
Robots are people too! Bam! That’s it!
Each human should be assigned a robot in which the robot does the hard, monotonous work and pays the taxes and the human just spends all day contemplating life and writing poetry about it.
It’s a pipe-dream for sure; one that corporations would desire to avoid at all cost. Do you think i’m the only one with this dream? No-o-o-o! There’s a Chinese kid in China soldering your future smart phone in some factory with no windows and no hoodies and — believe me — he’s got this dream! But he’s trapped at the moment. That’s right, he is. Right now he’s the robot replacement and it wasn’t very expensive either to set this up; all he had to do was give up his soul!
Bascly, the president is saying "remove me"
He wants to be removed, and he's just strange about it.
He's trying to get himself removed.
Dr. K says it's no issue, then some say it's an issue but we can solve it by big spending on infrastructure, then some say we can solve it by teaching people learning skills. Are you all God that can dictate what and how people do in the dire situation? No wonder NY times becomes the paper for the elite, out of touch with the normal people!
Paul, right, as is usual.
What is so surprising that Bernie gave a pragmatic answer to the automation question?
Bernie has a documented record of pragmatism going back to his days as the mayor of Burlington, where he worked across the aisle to improve the economy and got the potholes filled.
It’s terribly disappointing to read all of the pejorative comments in the NYT whenever Bernie is mentioned.
It’s condescending, it’s arrogant, and this lack of respect for our fellow citizens is eroding any sense of community, and undermining our very nation.
Mr. Krugman might feel very smug that he’s so much smarter than Bernie and the dolts ignorant enough to support him, but in fact Mr. Krugman is really just a genteel version of those Trump supporters in their MAGA hats roaring with approval while Trump speaks in degrading tones about the Democrats.
Mr. Trump knows how to throw the red meat to his supporters, and in Mr. Krugman’s own subtler way, his readers probably delight in knowing that there’s sure to be a little red meat interspersed in Krugman’s columns, even though the NYT readership would be aghast at any comparison with the coarse and vulgar Trumpians.
The connecting link is the contempt. Trump supporters express unseemly glee over their contempt, while Mr. Krugman expresses his contempt in the manner acceptable to people of good breeding.
But contempt is always corrosive, no matter how genteel the delivery.
1
How about building more schhols. hospitals, courthouses, libraries and concert halls and hiring more people to work in them? Or will robots do all these rasks?
1
Sabots: A French version of wooden shoes.
Sabotage: Throwing wooden shoes into the machinery of power looms to jam them.
People have claimed automation will cost jobs since Parry and Blake's "Dark Satanic Mills".
These machines MAKE jobs. Consider, highly ironically, the following:
In WWII, a young, wounded Soviet soldier wondered at the ineptness than had one rifle for every 3 soldiers in battle "One man falls, another picks up his rifle". Meanwhile the Nazi has a light, multiple function weapon that could fire singly or rapidly. So the young soldier swore to develop a weapon to protect his motherland that:
1) Could be cheaply, quickly and easily mass produced
2) Was light enough to be easily carried by a foot soldier, with plenty of ammo
3) Would tolerate incredible amounts of abuse due to looser tolerances than were usual
4) Could be operated and maintained by the youngest, most inexperienced, and ignorant Soviet recruit even under battle conditions
5) Could fire singly or pour lots of rounds down-range quickly
6) Used a common and easily manufactured ammo.
He deliberately did NOT sacrifice any of the above for accuracy--it was not to be sniper's rifle. His name was Dmitry Kalashnikov and the result was the infamous and ubiquitous AK-47 assault rifle.
Kalashnikov argued that HIS goal was protecting his nation and that his design became the favorite of terrorists and insurgents was never his intention.
1
Krugman is so old skool it's embarrassing. Yang has the most direct and cogent argument to date. Paul please do some research the ascendence of G, FB, A, Apple is predicated on automation at every level. GM 180K employees 50B market cap, Amazon 650K, 850B, Google 100K, 870B.
Plain MATH
1
One could debate the merits of 'priority' related to issues the Democrats should be talking about, but Krugman (an Economist, not a technologist) is putting himself out on a limb trying to predict a future where technology is out-pacing our (human) ability to comprehend its implications. Our politicians did not have a plan when NAFTA sucked 60,000 factories out of the country or closed due to competition. Why do we think the government is prepared for the effects and implications on the workforce of AI-driven automation?
Yay Mr Krugman, Laureate!! Good to see someone willing to think critically
He wants to be removed.
I see the notions.
Don’t you just want to talk in a way you are familiar with? How much better do you know about fourth industrial revolution than Biden?
Just do what you always do. Where is recession? How long should we wait more until we see the recession which you say it is coming every time stock market goes down. I recommend you to explain why your forecast was wrong instead of talking about any random topic you don’t know better than average people when stock market goes up.
Well, I guess we’ll see.
He's intentionally trying to get removed! I'm certain of it.
When was the last time a McKinsey consultant showed up at the doorstep of an American CEO and said “I’m here to help you figure out how to hire more American workers.”? I’ll take Bernie Sanders on night patrol with me before I’d take some Harvard MBA.
Nothing to gain by ignoring robots.
1
What are the real issues?
Good grief. Can he be more oblivious to the world? He admits that we have fewer farm workers and manufacturing jobs... and that is exactly why the Visionary Dr. Rev. Martin Luther King Jr. said we need the Guaranteed Minimum Income to eradicate poverty and make sure that all of our citizens will be able to continue exercising their freedoms by not being cut out of the economy. This writer seems to be the exact opposite of a Visionary -- maybe a blind man, but I don't want to insult the blind, to have all the facts and data in front of him after every single thing MLK predicted came to pass, and still to try to persuade people that it didn't happen. WOW!
If you can't see the structural defects of money, sucking money always into the big cities and back towards the wealthy... If you can't see the most pressing problems of our times as being a direct result of following GDP off a cliff, then you really have no business writing articles. You are wasting people's time.
"[T]he perhaps surprising exception of Bernie Sanders" - damning with faint praise? Apparently Mr Krugman still has trouble saying positive things about Bernie. Remember the last election cycle? Sample Krugman NYT quote from 4/8/2016 back when he was supporting the second most hated politician in the country: "Bernie is becoming a Bernie Bro."
I recall as a child in the 60s that my future was described as having to work only 20 hours a week because the machines would do the heavy lifting, and I would have plenty of free time to raise kids and explore my potential. Instead, I live in a country that works itself to death and farms out the raising of our kids. And free time? We save that up for retirement, if we don't die of stress related diseases in the mean time.
1
It is wonderful that you alluded to "Player Piano"! The worries about robots are a bit ahead of their time, and no one knows for certain what will happen, but it is worth noting that computer technology, by advancing logistics, is a very important factor in enabling the globalization revolution that has sent all the manufacturing jobs overseas. That is not "Player Piano," of course, but it is fully inside the high tech ballpark. It is part of the same thing. The global supply chain would not be possible without the little PC and Mac robots on our desks.
It’s partly economic justice in the answers from progressive leaders. It’s also recognition that the government has to define which practical science to commercialize for which public good, and support that coherently. Hiring for a 21st century world would no longer be left to the magic beneficence of private markets.
Maybe the progressives gave better answers because they think governments have to take charge, even own, technology innovation.
The problem isn't the robots. The problem is who owns the robot. If the productivity gains of automation flow predominantly to the properties there's a positive feedback loop that concentrates wealth.
The alternative extreme is where the robots do all the work, leaving higher pursuits to the humans (you can decide for yourself whether that's a Star Trekish utopia or Wall-e hell).
And then, there's something in the middle, in which "intelligent" systems augment human beings, and the economic are reasonably regulated against the development of black holes of wealth.
PS It doesn't take automation to create black holes of wealth, but we sure have seen the development of some of them over the past 20 years.
110
@Eben Am I right about this ?
It takes good government to maintain a balance between democracy and capitalism. The more inequality, the worse the government. Unfortunately, capitalism is winning now, so it is plundering and destroying the earth for profit. Climate devastation will be the mother of all market failures. We probably are destroying the planet for the kids.
http://gopiswrong.net/government.htm
19
@Robert Vogel Capitalism, IMHO, can be good, if it is appropriate checked. However, some of its underlying assumptions, of rationality and unlimited growth, are simply not fully true.
Kahneman and Twersky's research demonstrates systematic cognitive flaws. And anyone watch the overgrowth of micro-organisms in a petri dish, know where the limits of a closed system that's not managed with intelligence leads.
Unfortunately, our dreams of getting out of the closed system of the earth's biosphere is very unlikely, at least for our species.
The physics of money, under unregulated capitalism, creates black holes of wealth (and power), that is leading to disaster.
On the other hand, no one can deny that it has also made the life of an average human being (even in the worse of places that our President so ungenerously calls "shitholes") is better than that of our ancestors.
10
Truth is, no one, not even great Nobel economics prize winners, know truly and well where automation is going to push us. To corporations, cheaper is almost always better.
The immediate problem is that corporations treat most employees as disposable parts to be cast aside, while some, a much smaller number, are treated as superstars and paid anything the enterprise can support. Barrels of money for the few, penury, and debt, for the rest.
We have traditionally viewed poverty as a kind of unfortunate accident when, in fact, it is a condition imposed on people, one that is very difficult to escape. In a society where owning things is itself rewarded financially, those who don't control assets are then punished for their condition.
Most of us spend more than one third of our lives transferring money to the wealthy through interest payments on a house or apartment, interest that usually adds up to about 2/3s of the stated purchase price. A $400,000. house actually costs close to one million dollars over the life of a mortgage.
We have to find some way to share the vast wealth of this nation with more citizens. In recent times, the US has spent 3 to 4 trillion dollars on wars. What could we have done if most of that money had stayed home?
1
This century has all been about micromanaging employees from a distant corporate office using tracking systems with minimal local responsibility. It promotes no local input, no employee engagement or new ideas and limited opportunity for worker growth or new knowledge. It has been about cutting costs, ignores employee training and limits productivity growth. Continual education needs to be reintroduced, engaging and encouraging employees to think instead of just doing a task and providing opportunity along with living wages. American corporations need to rely less on centralization and more on their own great employees.
UBI and the automation question get at another question about what the future looks like and what individuals can expect as the inputs to their right to the pursuit of happiness. In a world where capital can create an increasingly huge amount of value due a to a decreasing marginal cost of production- and brilliant thinking and doing machines - do we accept that whoever was rich enough to control them is our new rightful aristocracy and everyone else is..not?
Or are we entering an era where politicians promising us that we will be able to work at jobs all day for our whole lives (and in so doing not have to go broke or die from treatable diseases or unexpected expenses that might hit us) is not appealing enough.
Our pursuit of a dignified existence in an advanced society might be reaching a point where we are no longer enamored of the often medieval-feeling system within which we work, age (if we're lucky), and die. How long can paranoia that there isn't enough to go around and the riches of an infinite digital walmart of 10 hour movies and cheap stuff from china distract us from this?
I've been a fan of Dr. Krugman for decades. But this time, he's wrong.
It IS different this time. Even before Uber and Lyft and Amazon Prime, we had about 7 million people directly employed as drivers. We are on the cusp of going completely self driving cars and trucks. Have you seen OTTO, the driverless semi truck? The big carriers are investing millions to be rid of their drivers.
What will 7 million drivers do? Become CNA's for our generation?
And no, most people cannot become AI engineers. Increasingly, the jobs that remain will take intelligence and education, to the levels that most people cannot attain.
2
I disagreed with Krugman as I agreed with Yang during the debate, who echoed my worries about automation seen at the grocery store, toll booth, and in views of some factories seen on tv and much more. I must say after reading the column I am more convinced that it is as much a "sort of escapist fantasy for centrists who don't want to confront truly hard questions." Yes, the real issues are present and need addressing. Thanks, Krugman.
1
...said the economist who thought the impact of the internet would be on par with the fax machine.
1
Krugman is right, and it's unfortunate so many of his readers are buying into the Luddite fantasy, which has reared its ugly head dozens of times over the past three centuries. Yes, AI is disruptive. But disruptive technology always creates new needs, businesses, and jobs that didn't exist before. We aren't able to contemplate what developments are going to come about from AI in this direction, but they will come about.
Dr Krugman,
Wondering what you thoughts are on this years Nobel prize in economics. If it can feed into an option piece even better. Us common folk need to understand the implications of good economic theories.
In the year 1900 the largest corporation in the world was a railroad. In 1950 it was an automobile company. In 2000 it was a computer company. Things change.
America now has a huge number of people employed in healthcare administration that add nothing of value whatsoever to pour economy.
Untold thousands of hospital billing clerks, pharmacy benefit managers and pharmaceutical companies attempt to bilk the system, whilst myriads of health insurance clerks try to avoid paying these bills by negotiating discounts and denying care before deciding how much profit to tell brokers to gyp from employers who pay for all this transfer of wealth from the poor to the wealthy. None of these people or jobs produce anything at all. no productivity whatsoever. all that cost is stripped out of companies and could have been invested in all sorts of great things. But no, we gave it to healthcare administration.
I, for one, welcome our new tech overlords.
1
Utopian views of the future call for we humans to relax in luxury, our every need tended to. Who's going to do the tending? Machines.
We're not afraid of the machines. We're afraid of not having money, because money = oxygen in our society. If we do not transition away from money at the same time that we transition to machines, we are in trouble.
With boundless respect for the Professor I must disagree. His is a jobs creation version of the deus ex machina defense.
Evangelical and Republican sorts will argue we should develop and consume fossil fuels as fast as possible. When they are depleted technology as with kerosene replacing whale oil shall provide new alternatives. Others simply argue God will provide.
Ponder it this way. What will life be like when all the goods and services needed for every living human can be provided by one person pushing one button once a day?
Sure that will not happen for a long time and we will have new opportunities arise for some time to come. But the coming future will look very different from the present.
Jobs are being squeezed out by technological progress. There will be a time when new jobs are few and far between. That is as obvious as biological evolution itself. How we might deal with it is hard to fathom at this point. But we best be thinking about it.
Railroads played a large role in the development of the United States from the industrial revolution in the North-east (1810–1850) to the settlement of the West (1850–1890).
By 1880 the nation had 17,800 freight locomotives carrying 23,600 tons of freight, and 22,200 passenger locomotives. The U.S. railroad industry was the nation's largest employer outside of the agricultural sector. The effects of the American railways on rapid industrial growth were many, including the opening of hundreds of millions of acres of very good farm land ready for mechanization, lower costs for food and all goods, a huge national sales market, the creation of a culture of engineering excellence, and the creation of the modern system of management. During the post-World War II boom many railroads were driven out of business due to competition from airlines and Interstate highways. The rise of the automobile led to the end of passenger train service on most railroads. Trucking businesses had become major competitors by the 1930s with the advent of improved paved roads, and after the war they expanded their operations as the Interstate highway network grew, and acquired increased market share of freight business. Computerization and improved equipment steadily reduced employment, which peaked at 2.1 million in 1920, falling to 1.2 million in 1950 and 215,000 in 2010. (WikiPedia)
The country and economy didn't die with all of those lost jobs.
1
Productivity "rose less than 1/2 as much during 2007-2018" as it did during the prior 11 years. "Which makes you wonder what Andrew Yang is talking about?" Well...yes, and it also makes you wonder why the rate of productivity growth is not faster.
Sometimes I wonder if Andrew Yang is a "solution" in search of a problem. That said, I certainly think it's possible for automation, if it doesn't replace workers, to make different demands on them. In that respect, "automation" might need to be unbundled conceptually. How much is job replacement? How much is an effect on the necessary skill sets?
First time I have ever found Krugman naive. The recent wave of automated improvements ( and demands) in medicine are destroying a profession. People really do not want automated 18 wheelers. What are people for. However, since it is Paul Krugman, maybe he is right, yet, I think we ignoring the warnings of Andrew Yang and Yuval Noah Harrari at our peril.
Paul Krugman says here that unemployment is so low currently because of low interest rates. What about the role of changing demographics? A much smaller cohort now reaching the job seeking age compared to the massive numbers of boomers retiring must have some effect on shrinking unemployment. I’ve read FED papers claIming that demographics alone has dragged down GDP growth about 1% from the norm experienced since WWII. It just seems that the role of demographics is often overlooked amongst economists.
Henry Hazlitt (of "Economics in One Lesson" fame) explained why technology has historically been a net job creator:
1. People have to design, produce, and market the machines. This aspect creates jobs.
2. When a firm buys those machines, it does so to improve its profits. It will cut some workers, replacing labor with capital, lowering its prices while expanding output. Other firms in the industry will also buy those machines to stay competitive. This aspect costs jobs in that industry.
3. The lower prices mean consumers have more money to spend on goods in other industries, creating jobs in those other industries.
Historically, #1 and #3 outweigh #2.
Democrats should stick to the basic message: We'll expand access to healthcare and higher education / trade school, funded by tax hikes on the rich. The ACA showed that a big government program can cover 20 million people without increasing the deficit.
If we can pry a few bucks away from the 1% at the top, and pay real living wages, we could also reduce the 40 hour work week to 35 hours, or maybe 30 hours, and give the increase in productivity gains to the workers in the form of more leisure time, longer vacation time, sick leave, family leave and the like. That would also keep more people employed, and generally happier with their employment.
4
Very disappointed in Paul Krugman's column today. While I get the gist of his comment, that there are many issues to solve at hand before automation creates a permanent unemployed and under employed caste, we need to begin discussions of the very real possibility of massive disruption to our social structure.
Never before, never, have we had the possibility of generating all the goods and services that we need for humanity to survive, using a fraction of the workforce. If the distribution chains and political divisions are improved and repaired, today we probably only need about 50% of our laborers. And accelerating. Examples of this abound. What then?
This dramatic shift in the social paradigm is scary and exciting. Scary, because idleness is not an option, and will create massive upheavals. Promising because of the exciting possibility of working for fulfillment and not for survival.
Just like with pollution and the environment, the time to talk about automation is now. Sorry Paul. We are not Luddites. Neither should you be.
1
How meaningful and fulfilling are these jobs that can be done by robots. Do you really want to stand at the supermarket checkout counter or put together widgets on an assembly line for the rest of your life?
Even coding has become automated, some good IDEs fill in the rest of the codes for the developers, all they need is the architecture and logical flow.
Over the last 10 years, over 2.2 million jobs were given to H1Bs from India, let alone the outsourcing e.g. Cisco to India.
So yeah, domestic spending might improve the job situation as long as the jobs stay in the country, but what sort of domestic investments, and what infrastructure projects?
We need a global discussion on these issues.
Sorry to disagree.
Denying the possibility that machines – robots – one day soon will do everything better, faster, and cheaper than even child labor or student interns working 15 hour-shifts at Foxconn is like saying that Chinese fireworks could not possibly be improved upon and get a bunch of robots to land on Mars.
Mr. Yang has limited his contribution to the discussion, but his ideas are no less radical than Bernie’s and El’s.
True: No one knows what the future holds. But we should entertain the thought that the future will be radically different from – say – going from family farms to industrial farming.
Mr. Yang’s proposal is universal income.
I am for testing it in places that desperately need it. We all know where they are.
People don’t understand the mechanics of quantitative easing, but they get pocketbook easing.
There is still the issue of getting Bernie’s, El’s, and Andy’s ideas past the GOP guardians of free market corporate subsidies. The challenge will be not to tire of the discussion before the future arrives.
Another insightful article by Prof. K.
Many comment below that its the pace of change this time. Yeah, this time, its different. Except when it isn't ... which is every time.
I continually enjoy reading in the comments section, the self-appointed experts (sarcasm; you're all very much not), with their anecdotal stories "tsk tsk tsking" a Nobel Laureate.
Do keep it up, its good entertainment.
Mr. Krugman has a rosy view that an automated future will bring new job opportunities in some yet-to-be-imagined field of work.
I see it differently: Capitalism is an economic system to maximize the return of profit on investment, and human beings (in fact, life in general) are not a priority to that system. If it is more profitable to make the planet a wasteland, capital will move in that direction.
Example: Mark Carney, the governor of the Bank of England, is quoted in The Guardian (16 October 2019) stating that "if you price the capital markets, and I'm not giving you a precise figure, that all of the assets are probably north of 4 degrees C for the capital markets as a whole." In other words, the flow of capital is toward investments which will lead to global warming of at least 4 degrees C (about 7 degrees F). Nations may negotiate all sorts of climate treaties with good intentions, but they won't stop capitalism's forward march.
If that's how capitalism treats our environment, why should anyone expect capitalism to be kinder to us humans? Maybe it's time for humans to recognize that capitalism itself is the ultimate killer robot, and that AI is just making it smarter and stronger and faster.
1
Speaking of U.S. jobs lost to automation, what about those, potentially, lost to a (eventual, if ever) implementation of a "Medicare for All" scheme? I'm referring to the ca. 2.7 million people employed in the health care insurance industry. Not that automation would be the direct cause but it's something else to add into the mix. I wish that one of these debate moderators would add that question and see if Warren and Sanders have a cogent answer.
I work in the insurance industry, and I’d be happy to lose my job (and pay more taxes) for a single-payer system. Furthermore, the new system wouldn’t run itself—people would be needed to help run it. If we aren’t willing to seek new opportunities in our own careers in order to better the country for all, we ought to be ashamed of ourselves.
I have been in heavy manufacturing and did automate some processes, notably CNC metal cutting, which doubled or more our production, but also eliminated a lot of really bad injuries (and chronic hearing loss for the younger guys - to late for me).
However, most tasks work better if people do what they do well and machines do what they do well, working together, redesigning processes and no even products to optimize the whole system. One of the reasons productivity hasn't increased is precisely because management and the people who do automation don't understand this and are just trying to replace, not enhance, people.
Humans are also good at the exceptions. I don't know how an autonomous Uber will fix a flat or clean out the back seat after transporting a group of teenagers who literally can't hold their liquor.
Autonomous ships are another idea that I wonder about, especially considering some of the problems with cargoes (a container bananas suddenly ripened and had to be gotten off before it ripened all the rest, a bunch of full acetylene cylinders came loose and were rolling all over the deck, various fires in containers), not to mention interested failures in the engine room.
Jobs will change, but AI so far hasn't gotten as smart as a hamster, so people will still be needed.
Yes, there are larger issues right now, but the robots ARE coming. Denial of the high impact of this is a rabbit hole of it's own.
1
Krugman: (Yang is) "offering an inadequate solution"
Krugman: "The persistent weakness... is about inadequate private spending."
Sure Paul, there's no way that giving every citizen $1,000 a month would increase private spending.
1
This is a painfully weak, narrow-minded piece. "This time is different" not because of the current pace of technological change, but because of the current and predicted nature of technological change. We're rapidly moving into an era wherein highly technical white collar jobs -- like surgeon, radiologist, and lawyer -- will be replaced by machines. And these are not supplementary technologies, that make the surgeon, radiologist or lawyer more productive: they're complete replacements.
The author is applying old paradigms to a radically new environment, which is misleading and comforting. People who are actually familiar with the trajectory of AI are much less sanguine, and that should concern us.
1
As a machine learning engineer, I have a few things to say on automation:
1. Right now, artificial intelligence is rubbish. It's glorified differential calculus that's only successful because supercomputing is just so cheap and plentiful.
2. In the next thirty years (realistically), there are going to be breakthroughs in AI and robotics that will have enormous ramifications on the relationship between human beings and the concept of "work."
Sure, jobs have been displaced by automation before, but the technology that did the displacing also created jobs. But this IS different. Previous surges in automation have replaced people's arms and legs as a means of getting things done. The hoe was replaced by the tractor. A human still drives the tractor. A human's effort was multiplied.
This time, however, automation is going after the human brain. This is substitution, not multiplication. We have robotic arms. We have electric wheels. We are now building the brains.
Also, we're going to need to start distinguishing between the economy and the society. The economy will be fine. The GDP will only get better. Automation will drive non-automated job wages down as there will be more human supply for an ever shrinking human demand yet there's a threshold of cost at which it becomes worth it to automate (human labor is only valuable if it's cheaper than the automated equivalent). As a result, income inequality will increase drastically.
Wait a minute.... does this sound familiar?
5
All very good points, Dr. Krugman. Just possibly Elizabeth Warren will pay attention; unlike most prominent politicians, she's shown signs of learning something other than tactics from experience and new information. Otherwise, I think that the people going on about automation are impervious to facts.
1
Normally, I find myself in agreement with Dr. Krugman’s analysis. But this time, I find his argument is rather poor.
First, he straw-mans Yang’s position. Yang did not say his UBI would solve the problem of jobs vanishing, merely significantly soften its effects while providing a large boost to the economy, to give the country more time to deal with a large number blue-collar jobs vanishing. Yang has also pointed out, bluntly, that economically insecure people are much less likely to vote for policies which are perceived to help Those People (Krugman’s apt phrase)... which comprise a large part of the progressive agenda. We want our agenda to happen, right?
Second, economists have been quite oblivious (as Krugman has observed) about some effects of automation and free trade on workers. Yes, the economies of the countries involved grew, but the benefits didn’t flow to the people displaced from their jobs and homes. There were winners, and there were losers. I don’t think the losers felt any better that somewhere else, in a city far, far away, people benefitted immensely from a trade policy that unemployed them.
Now, we may be on the cusp of another enormous wave of A.I. powered automation destroying jobs. There are many billions flowing into replacing truck drivers, an industry that employs five million without college degrees in the US. By the time you see that show up in the productivity figures, it will be too late to come up with a good policy!
1
largely true, with one suggested correction. Unemployment is really low by historical standards, because labor force participation has never recovered after 2008. With that recovery, the unemployment rate would have been ~7.5%. It could be that the US is entering a period of relative labor shortage and more automation is required to cope with it, particularly if we limit immigration? Double digit unemployment may be the thing of the past, at least until some severe economic shock arrives. It is the quality and pay for the jobs available that may be the problem. After all, Ohio auto workers are not exactly complaining that there are not enough McDonalds or Walmart jobs for them to retrain into.
As for Andrew Young, he is kind of full of himself as most techies are. This is what being showered with stock market monopoly money does to your brain.A new breed of Masters of the Universe (he thinks). Airplanes were fully capable of automated operations for the last twenty years. They are not being automated for safety reasons. The same will go for self driving cars and trucks. Driving is a much, much harder problem to solve than flying. The air around an aircraft is mostly empty, compared to auto traffic aircraft traffic is sparse, fairly easy to regulate and highly predictable. Self driving cars will be permanently on the cusp of a commercial breakthrough for the next 20 years, and then they will simply be banned on the grounds of liability concerns.
1
My business, like many, would not exist without the digital transformation we've seen. Clients I have never met in person, would not be clients--and I run a service business, traditionally based on in-person interaction. What I don't have enough of, is highly skilled thinkers--people who can solve problems on their own, think outside the box and deliver genuine insights. Automation has not displaced those people. We just have a severe lack of enough of them, and that is where investment needs to come in. Education geared toward creating problem solvers is everything in the new economy, and it is truly lacking in that area. More often than not, recent college graduates are grossly self-absorbed in social media and grossly ill-equipped to make meaningful contributions in the modern workforce. Finding those that are ready is like searching for a needle in a haystack. There is abundance of excellent work, high-paying work to do that doesn't depend on automation--but no one to do it.
1
This whole idea that machines will do the thinking for us has not panned out, and it won't. Sure, a large fraction of what we do can be done by machines, but there will always be humans taking care of things that we have not yet told machines how to do. Automation, done right, lets the mundane get done while we take care of less structured things.
1
In my profession, retired librarian, automation didn’t cause a loss of jobs, but gave us new opportunities to provide better in-depth services. Also, automated machines need to be repaired, reconfigured, explained, etc. All of which takes time and skill. Our education system always seems to be playing catch-up with what is going on in the real world. Monies need to be put into it, and many more skilled educators are needed. When I worked, I was constantly taking classes, learning and expanding my skill sets. Employers may not offer that opportunity but it needs to be made available. Education needs to be easily and affordably available at any age and meet today’s challenges as well as serve as a bridge to culture and the past.
Regarding consumer spending. If we are going to become environmentally more secure, we have to rethink how we spend and stop consuming stuff that will be thrown away or break within a short period of time. We have to figure out a way to make our refuse heaps smaller. This means recreating our economic structure also. More repairs, more skilled workers, more thought into what we purchase. The consumerism model of the past should and could become obsolete with more of our monies spent on learning. That means less plastic and more teachers.
1
"Player Piano" is a very chilling novel to read because it is so very plausible. Historically, we have theorized that automation increases the productivity of every person-hour and allows us to make things and provide services more cheaply, thus leading to increased economic growth that will generate new jobs, of some kind,somewhere, to replace those lost to automation.
Some will argue this has worked for a while. It must be, as the unemployment rate is at record lows, right? But the labor force participation rate isn't good. And the replacement jobs don't go to the same people in the same place. (Should we have paid to move a big chunk of the population of Flint to Cupertino and Redmond, thirty years ago?) It takes far fewer worker hours to build a car than it did in the 70's. But cars are not getting cheaper. Middle class consumers are now taking out 6, 7 or 8 year loans to buy new cars, instead of 4 or 5. Contrast this with the 1920's where increased efficiency steadily drove down the price of a new Ford.
Oh and by the way, the "we will create other jobs" argument presupposes that the economic benefits of automation are in fact invested in job-intensive growth, rather than being vacuumed up by executives and shareholders in bonuses, dividends and share buybacks. Hmmm.
Ultimately, there is an inherent perpetual-motion-machine, musical-chairs magic to the theory that there will always be a replacement job.
3
It's not a question of avoiding the cliches about automation and jobs and education in America. It's a question of what we invest in and what direction America goes in the future. For too long there hasn't been any investment or interest in 99% of us. We've been viewed as cash cows for decades. We've subsidized the richest people and corporations with tax cuts at our expense in the belief that doing so will create jobs that pay well, encourage investment and good behavior. What we've gotten in return is anything but that.
Technology is not the answer but it's here and it's being pushed out as the answer to every ill imaginable. We are social beings no matter how introverted some of us are. I choose to use the human check out lines in the grocery and in other places. The self checkouts are too temperamental and not intuitive enough. When I have to go through a computer menu on the phone I select speaking to a person. Why? Because the automated choices don't usually help. If I'm calling it's not about my account number or something simple. It requires a human at the other end.
We are using technology in ways that are hurting us rather than helping us. Yes, there are jobs that ought to be automated: the dangerous ones that don't need people, the jobs that can put people to sleep because of their repetitive nature. As long as we live in a society that requires work for money we have a duty to ensure that anyone who wants a job can have one.
3
Do companies with robots still pay into Social Security and Medicare? Are “value” taxes the way to go? Inquiring mind wants to know.
1
Respectfully, this current issue isn't necessarily that there are too few jobs, it's that the type of jobs have changed away from manufacturing, which was absolutely caused by automation, and which absolutely contributed to job loss in the midwest and Trump's victory.
Also: while automation is a good reason to have UBI, it's not the only reason! Ending extreme poverty, giving people freedom to leave abusive jobs, giving people freedom to leave abusive relationships, making sure everyone ALWAYS has a little money coming in no matter what (something that rich folks like Krugman discount because they've never been poor), reducing inequality, putting money back into local communities and encouraging people to stay in rural areas...you can go on and on.
Finally - we can afford it! Why should billionaires get government subsidies but not normal Americans? I really think Yang and his ideas deserve a fair shot.
I am not surprised that Bernie Sanders had the best answer on the issue.
So many people are used to saying that since he is a self-described socialist he doesn't understand society or its problems, as they keep saying socialism doesn't work.
However, his political record, starting from his first term as mayor of Burlington, VT, suggests that he remains a socialist specifically *because* he has a good grasp of the problems and how to fix them well.
1
Prof Krugman mentions Player Piano, based on the Great Depression, which was blamed on automation. Factories didn't need as many workers, and without those workers employed, there was less demand for what the factories produced, so the factories reduced production and workforce.
But Player Piano must be placed against the Jetsons, where everything was automated but everyone had a job that paid enough for a nice apartment, a flying car, and a robot.
The difference is explained by the IS-LM economic model.
Back in the '50s, earned income had a top marginal rate of 50%, and unearned income 91%, and the IRS said executives who paid themselves more than 5 times as much as the average worker were getting unearned income, and taxed it all away. So unemployment was very low, and most jobs paid about 20% as much as the incomes of the richest Americans.
We need to go back to that.
1
Robots are not the enemy: efficiency is. Robots are a product of the insatiable quest for efficiency... but so is the idea of optimized scheduling of employees to increase profits and reduce the full-time work force; so is the privatization of public sector jobs; so is the outsourcing of work to lower-wage workers; so is the creation of factory farming; so is the deregulation of environmental and safety regulations, and so is the deregulation of ANYTHING that erodes the bottom line and doesn't "add value" which is defined in three month windows. Efficiency is Holy Grail for both Democrats (think "Reinventing Government") and the GOP (think about any ideas they come up with that fatten the wallets of the plutocrats). Robots eliminate jobs... efficiency eliminates the joy of working.
1
Well, just because the automation didn't replaced workers in 50s, doesn't mean it won't replace them all in 2050. After all, automation in 50s led to higher salaries, while automation in 70 till now, led to stagnated wages. Clearly, something changed. It is not very honest to compare growth productivity 2007-2018 with 11 years before, considering the Great recession in 2008, but the wages stagnated way before 2007. Could it be that automation is moving people from higher paying jobs to low paying jobs, rather than replace them totally out of job market? It may be different problem, but it is still a problem that need to deal with.
2
@yulia - if you've read PK for any time at all, you should know he's interesting to read but cherry picks numbers and dates at will to "prove" his points.
As an engineer in manufacturing who has suffered through the automation changes of the past 40 years I can tell you that Warren is mostly correct. But there is more to it: MBA types see technology and think, "Great I can get rid of all those pesky engineers, foremen, and operators." (MBA programs are largely a failure because they reject process.) When you flush 80% of your workforce out the door you lose vital expertise. Add globalization, and now MBAs think they can move all those jobs overseas.
Other countries are not as foolish.
Recently, our plant suffered a rupture disc failure that kept our plant down for a week. That disc was made in Mexico: there was a time it would have been made in America. Instead, we used one made in Europe and our failure issues went away. I've seen the same problem over and over again as corporations kill off the middle class in the US to save a few bucks because they think no one will notice poor quality: consumers are watching.
4
I spent the last 18 years of my work life as a tech worker going into firm after firm automating business systems with web based technology. We'd come into a place and a year later hundreds of people would be put out of work. It's devastating to see.
I believe that we have not developed the social and moral systems to protect us from the awesome power of these new technologies. Yang does not have the answer, but we do need to start considering the problems which will arrive as we increasingly replace natural systems with digital systems, human intelligence with AI.
3
Who else remembers when Dr Krugman predicted the stock market would NEVER AGAIN get to the level it was on election day 2016.
Our future would be pure decline forever going forward because the voters didn't like Hillary.
Look to San Francisco as the near future. Elites living like royalty and everyone else who isn't homeless struggling to get by.
AI makes one worker 1000x more productive than another. Is that the way the spoils should be split too?
3
What you fail to realize, that unlike in the past, at some point, some say in about twenty years, the machines will be smarter than us. At that point they will be able to do what we do better. It’s as if we are engineering our own extinction.
3
This all seems to boil down to problems of human nature. People take advantage of other people. People want to feel better than other people. The rich can never get enough. We all want more, faster and cheaper. It's hardwired into our DNA.
So we probably can't solve these problems. But the onslaught of automation will solve them for us.
Eventually, there won't be any more people.
3
Dr. Krugman should heed his own advice about looking at real issues and using his wisdom to offer sage advice. If Andrew Yang is barking up the wrong tree with automation and a "freedom dividend" as PK claims, then I would ask the noted economist to examine the impact of aging baby boomers being heavily invested in market index funds for their retirement savings. I'm betting that a good portion of those commenting here are in that group. And while we lean left politically we vote like Republican's with our money. Dr Krugman: what will be the impact on the economy when the elders draw down those funds?
1
There are plenty of jobs- if we will just at long last repair our infrastructure- which includes building a new and modern power grid. We can't use robots for that- yet. And the alternative energy manufacturing industry is poised to grow big. But every new and old job must be brought in line to reflect a new minimum wage of at least 15.00 per hour- and jobs above this type of "starter" position must be supported with higher wages to reflect experience, training and education. Might not need a stipend, then.
As Paul Krugman correctly points out - the issue isn't jobs. We now have one of the lowest unemployment rates for the longest sustainable period since WWII. The issue is wages. The threat of substitution from foreign labor and machines puts a ceiling on the wages Americans will be paid. The two threats, however, work in opposite directions with respect to jobs. Automation allows jobs to stay in the U.S. that otherwise would have been sent overseas.
I’m afraid this piece ends up confusing the issue. One problem with discussing automation is that it gets confused with the more general topic of technology change.
As Krugman points out, it is a mistake to believe that we can base a jobs policy on getting back all those jobs that have gone away. There are many reasons why they’re not coming back. For one thing technology is fueling a longstanding movement from manufacturing to services.
That doesn’t mean there aren’t jobs, and Krugman (citing Sanders) is also right about the importance of growing the public sector to do the things that aren’t getting done.
The problem is that when people start talking about automation, narrowly defined, they can quickly jump from “I don’t see that many robots” to “it must be just trade”. That’s what happened with paper Warren was citing, and she’s not the only one.
The world is changing, and we’ll have to change with it. That’s fine. We can do it. The problem is not acknowledging change.
Just because it is an exaggeration to think that there will be no such thing as a job in a decade doesn’t meant that the idea is without merit. I consider the UBI or negative income tax to be a preferable alternative to the minimum wage, in which we as a society pay for a minimum standard of living for people rather than taxing the specific companies that happen to employ low skill labor. If, for whatever reason, wages are being depressed, as has been the case for decades, with massive wealth accumulation by the wealthiest people and little change for workers, a UBI is a method of addressing that. Whether automation is the cause of the wage depression or not is irrelevant.
1
Krugman, get your head out of the sand. The only reason we aren't talking about automation every single minute is because most of the jobs that are easiest to automate happen to be the types of jobs that America has lost over the past several decades to overseas competition. (Warren's point about the "trade problem" is accurate but irrelevant.) There is no question that automation is a larger and more immediate concern in those countries that now dominate sectors like manufacturing or call centers. These are the jobs that are the first to be lost to automation but they are not the last. There is no reason that automation and AI won't soon begin to take root in the sectors that continue to thrive in the US and that's a big deal. A very big deal.
The fact is that the human workhours it takes to produce the food a person needs to survive has gone drastically down and will likely continue to fall. The same can be said for the production of all the other things we need, clothes, shelter, transportation, etc. Ignoring the distribution problem for now, we as a society have to decide what we want to do with the fact that the number of workhours required to provide a person with a comfortable life, is and will continue to decrease.
We could retain the same definition of “comfortable life” and reduce the number of workhours that society needs from each person on average. Alternatively, we could “up-define” the concept of “a comfortable life” to include more or better things (as we have done so far). We could also decide that the definition of a society should be upgraded, so our spending of work hours goes more towards collective goods and services.
This is a major strategic decision that should be made before we can even begin to figure out how to change distributions of things in a fair way.
1
If a machine can perform a task at lower cost than a human being, the machine will get the job unless the human being is willing to work for less than the cost of installing, operating and maintaining the machine. Robots may not at this point be net displacers of jobs, but they clearly are having an effect on what competing humans are paid and thus contributing to inequality. Protectionism will not promote the return of middle-class incomes for factory workers. It will simply hasten automation and doing little or nothing to raise wages. What we need is surging demand for a whole new array of goods and services along with policy interventions to reduce the cost of labor so that workers can compete more easily with machines. Moving with urgency to a zero emissions economy would help generate the needed demand. Serious health care reform would reduce the cost of labor.
Kodak and Poloroid employed 130,000 workers. They are all gone, replaced by Instagram, that employs less than 100 people. Automation is more than "creative destruction". Its downright devestating and doesn't always lead to an equal or greater amount of new jobs. People who are devastated vote, often for someone who will destroy others to spread the pain so they can say, "Now you know what it feels like."
2
Yeah. Ask the 19th C weavers in Scotland about losing jobs to automation. The disappearance of jobs there brought Andrew Carnegie and family here. Andrew, having learned from experience, made a fortune by embracing and mastering every good tech innovation to come along. Nevertheless, I would gladly take Yang’s $1,000/month. We can’t all be as resourceful as Carnegie.
1
The rise of the computer has changed the workplace. When I started working as am early Boomer, a computer was something the size of a small building and was addressed with punch cards. The calculators I used were mechanical and resembled a large office typewriter. Once computers were reduced in size and given enough memory and power to control robotic machines, everything started to change.
I worked in a medical lab where technologists would carry out an assay by manually measuring out quantities of liquid with glass pipes. After perhaps several additions, the tube would be mixed and perhaps incubated at body temperature for a time for color development. The tube was then place, one at a time into an optical device that measured the intensity of color. The concentration of the substance being measured was then manually calculated, sometimes using a hand drawn chart.
Today, labs can have a moving belt where the specimen tubes move from analytical machines that do all the assays and feed the results into a patient record. There is normally little for the human to do other than removing the collection tube stopper and placing it on the conveyor. Obviously, the machines must be maintained by humans, but barring mishaps, the machines can run largely by themselves.
Technologists must have a degree in NY and often have a masters in science. Relatively highly educated people are being displaced by automation.
5
Blaming automation is dangerous.
Automation is not new, it's just been getting steadily better over the last century (and longer). Automation increases productivity so (generally) lowers prices on goods while raising quality and value, improving our standard of living. This is good.
Yes, the dumber jobs (e.g., inserting a gadget in a slot on an assembly line) are going away, and new needs (designing and building web sites, installing new communications gear) are appearing. Things like that have always changed, and they always will.
Yang is pandering to fears about automation, but from a big-picture view, those fears don't make sense.
2
Andrew Yang reminds me of old time science fiction radio plays, in which robots are brought into the workplace, where it is assumed the machines will do the work they are programed to do , however as time passes the machines become smart and turn around and dominate their old masters. The plays usually end with some one screaming to shut them down or in silence... the lister being lead to assume that the machines win. These plays still plug into a lingering fear of being marginalized at the work place and a loss of control over the social impact of modern technologies.
Economic inequality is very much a product of the way technology is utilized. In the past factory workers displaced by automation could enter an expanding service sector unfortunately also leading them to lower paying jobs, with the net result being low unemployment but little to no wage gains for the middle and lower working classes.
Today the service sectors are in danger of being supplanted by artificial intelligence that can imitate the human voice and judgement and may soon even have a face. It is easy to imagine businesses replacing there service sectors with such automation as machines don't need health care , time of for illness and don't go on strike. The issue may be only one recession away as businesses may conclude that survival will mean not just layoffs of factory workers but replacement of service workers with artificial intelligence. Yang may not be as far off base as some imagine.
Bravo! I was appalled as I watched that part of the debate, and glad someone called them out on it. You do go a bit easy here on Warren. I admire her, but her awkward “I want to see the data” hedge send more like a political dodge meant to avoid offending Yang fans than true confusion or humility.
1
I am not sure if the guaranteed income should be sold as a way to counter automation or if I should be sold as a way to reduce inequality. I think that all liberals would agree that we have an inequality issue in the united states. Making joining unions easier will help, but given the current nature of employment in the US I don’t expect it to solve the issue. A higher minimum wage will also help, but also will probably not solve the issue. Why shouldn’t we consider a guaranteed income as another tool to fight this problem.
2
My take on Yang: We, the world, right now have populations with more than they, as in we, can use. We need to back off the production - just cool it. Assess, plan, adjust, be smart, look ahead . . . and forgo excess. Fair distribution - now there's a challenge. Unfair distribution - picture a world according to Trump.
My take on Krugman: Almost always valuable and about right, but he missed it this time.
2
@micheal Lubell
I agree whole heartily, the key to the future is learning skills, I.e. lifelong learning. Skills and the ability to question technology as well as problem solve, will be the future to employment in the twenty-first century and beyond. The crucial step is for all students is to learn to think not regurgitate rote responses.
1
Consider, please, Dr. Krugman, that all automation is not the same. Automation for most of the world in the past was automation that could be understood by the workers assisted by it. My father, a brilliant mechanic, and inventor,could not make the switch to electronics. In 1960 at the age of 55 he threw in the towel saying he could fix anything he could see, but could not work with the intangible. He resigned from one of Americas largest producers of automation to repair old clocks. Automation has now gotten way beyond the comprehension of most people who depend on it. All manifestations of it come as an unexplained given, so that the person operator is, as the sophistication of the automation increases, reduced by comparable measure to automatic gesture and involvement, and that suggests a dumbing down, a numbing of human endeavor. I'm not suggesting you are wrong, Dr. K, but it just may be helpful to stop examining technology in its myriad forms and effects as a monolithic entity. Each application has its social, political, and psychological effect. Phones are removing us from traditional human contact. A driverless car will separated us from responsibility. There, you have it, the ideal life: drive around all day with your feet in the windshield and a phone in your hand. If Yang will get me 4,000 a month I'll vote for him, but his 1,000 dollars is for the birds. That's my wine and cheese budget alone.
4
Sorry Paul,
I thought you might have a deeper understanding. With corporations replacing human labor with automation, labor becomes less valuable and the owners gain the benefits of the trade, increasing income inequality. This is why more and more people will have a harder and harder time making enough money to afford what they need to survive.
6
We need a debate that gets to the root of the problems we face: gross wealth and income inequality. Automation, artificial intelligence, trade, private equity, deindustrialization, finance, immigration, lack of education, etc: each contributes to the economic dislocation that affects Americans by costing some their livelihoods. We can debate the net gains vs. losses for society, but they clearly can be devastating for large segments of the population, hence the growing number of people being left behind. Policies that that get the balance right on these issues are important to pursue. But more fundamentally, we need to grapple with gross inequality directly. I buy lots of stuff at Amazon, and sometimes I shop at Walmart. But a political economy that results in one man (Bezos) and one family (Waltons) owning more wealth than hundreds of millions of their fellow citizens is not sustainable. That is a system that works incredibly well for the few, not the many. Mr. Krugman and other economists, and policy makers who really want change, should turn their attention to why and how our laws and policies contribute to such results, and propose how to fix them. Spend more time trying to cure the disease, less time debating how to mitigate the symptoms.
3
Vonnegut's Player Piano was one of several dystopian sci-fi works warning of the consequences of automation. A lesser known work by Frederik Pohl, The Midas Plague, also published in 1952, presented a world where robots take over all manufacturing. But to keep the "supply and demand" system going requires continued consumption. This falls onto the backs of the poor who, in a bizzaro world twist, are required to consume what the robots produce and are exhausted by these lives of excess. The "rich" are allowed to consume what they want and live comfortably. This premise takes economic theory to an absurd extreme, but with rampant consumption required for a stable economy is it much different from today? Humans and robots need to find some kind of happy economic equilibrium.
1
Inadequate private spending? My confusion about the pace of automation, which A.I. Is bound to stimulate, is if there are no jobs, who buys the goods robots produce? It seems something systemic has to change. But what?
1
I don't think automation is THE problem, but I do think that accelerated depreciation has tipped the scales toward greater investment in automation, rather than in new hiring. Year one depreciation is especially financially beneficial to medium sized business manufacturers, who may intend to sell or merge their businesses in the next 2-5 years. Why hire 25 new employees when you can invest in automation with rapid capital depreciation? Accelerated automation is a greater economic force, than the predictable historic steady flow of automation that Dr. Krugman identifies. Economists and politicians must have a new vision to deal with accelerating automation's effects. It's here to stay. Yang gets it.
I'm not so worried that automation will eliminate jobs as Krugman often describes the workings of macroeconomics determine the level of employment. The problem seems to be that workers are effectively bidding against automated systems to determine pay levels. It's systems--not robots-- that are the problem because rather than a robot taking your jobs, the pattern seems to be computerization splitting the work into pieces that can be handled without great skills and at low pay.
Robotics makes sense only when it improves the product or service the corporation is selling. Robotic processes that make car doors that fit better and offer better protection in crashes sell cars. Robotic customer service call centers that can handle more calls, process customer requests faster and satisfy customers needs, please customers and generate more sales. A cheaper car door is just cheaper. An automated call center that handles call faster is just faster. Robotics is just a tool that can be used either effectively or ineffectively.
1
It is much more difficult to measure productivity change in the service sector than in sectors that produce "goods." Perhaps the perceived slowdown in productivity growth reflects the change in status of many low skilled workers from unemployed to low wage service workers, such as caregivers to the elderly frail.
Finally!
Finally a left-wing economist who admits that most of the millions of jobs lost in the US in the decade before the start of the Great Recession was the result of stupendously naive trade policies from the Bush and Clinton Administrations.
Finally a columnist who is prepared to fact-check the fact-checkers and call them out.
Finally a liberal with an intellectual pedigree who is prepared to admit that Warren and Sanders are the hard-headed realists in the room.
1
Agree with all. You did not mention Mr. Yang's plan to pay everyone $1000 per month by implementing the Value-Added Tax. The VAT is highly Regressive, as it is passed on to consumers, which means the poor pay higher percentages of their income in taxes than the rich.
1
There will never be a time when machines replace people. And their is so much demand for workers in fields never thought of before, demand for help for the elderly and disabled, demand for child care, demand for repair and replacement, demand for services. Its just an adjustment to a way of life never thought of before. The sooner we realize this huge need the better off we will be.
1
Of course there will be good paying jobs available for humans in the future. The problem is that they are likely to require ever higher levels of cognitive skill that will leave greater and greater numbers of humans behind as technology advances.
"...there’s plenty of work to do in America, and right to call for large-scale public investment..."
"It finds that a quarter of U.S. jobs will face “high exposure to automation over the next several decades.”
"When, in modern history, has something like that statement not been true?"
Sometimes I feel that we're all already deep down a dark rabbit hole. Automation, AI, manufacturing, the entire economy rests on a crumbling, outdated infrastructure. This infrastructure rests on a changing climate landscape. We have declining resources, increasing population, and are not facing the sacrifices we should be making to prepare and mitigate climate chaos. We have perpetual toxic nuclear waste sitting vulnerably in 'temporary' storage sites, and compromised air, water, and land quality. What could go right?
We should be teaching people lifelong learning practices, with an eye and hands on approach to the reality of our precarious situation.
I grieve for the grandkids.
2
I read "Player Piano" in the late 1950's. The vision in it that most impressed me at the time was that the people of "Illium" in the novel ( which was modeled on Schenectady NY) were divided into two classes; the company technocrats leading normal lives, and the much larger group of unemployed, who had lost almost everything.
Take a look around Schenectady today. See the effects of GE closing down manufacturing. "Player Piano" 's predictions look very relevant.
2
As long as we Americans continue to prioritize $15 lawn chairs over $15/hour living wage, we will continue to empower corporations to decimate the middle class ....... until inequality reaches a point where it measurably diminishes US demand (and such demand is not adequately replaced by demand by consumers in China, Vietnam, Brazil, India etc).
Then. Change is coming.
Or, we can wake up and start looking after our collective interests by restructuring how our businesses are structured. They are built to solely reward investors .. to the exclusion of workers. Insanity.
2
Paul, Please explain "it’s about inadequate private spending". I'd like to hear more about that. Is this just code for income inequality or is there really a crisis in private spending? Personal savings rates are up since the great recession (5-6%) but no where near the 8-10% we had from 1950-1985.
Very cool, but what is causing income inequality and were are the jobs? Automation and AI could replace a lot of people up the worker chain, which could depress wages, widening inequality. Our economy is growing slowly and that is probably because it is approaching a "steady state" that we see in physiology and closed systems. I expect that the income inequality that we see is because there is fewer high paying educated skill jobs than muscle jobs at factories. We have saturated the market for refrigerators and washing machines, now iPhones and Health care.
Where are those higher paying jobs today? Health Care. And we pay a big tax in high insurance premiums. If we cut costs to appropriate levels, what a massive recession we would have. And keeping us septuagenarians and octogenarians alive does give our children and grandchildren jobs but they pay dearly for them in their own premiums. Where else is there massive growth in our economy? and how long can HC support us? Just asking. And inequality just keeps marching on, its economics.
1
Why aren't we using all this automation to reduce the workweek, to give us more leisure time, to enjoy life? Are we eternally condemned to working like Elon Musk, 80 hours or more a week to build a better spaceship, build a better automobile, and so on?
What if the workweek was four days; more workers could be employed to build a better spaceship. They would also have a family life, take vacations, go to concerts, barbeque with friends.
4
By all means we should pursue solutions to the real problem of a slow economy.
As some candidates suggested, migrating jobs to green industries would be part of the solution to job losses; however, somewhere down the road isn't a shorter work week and higher hourly wages part of the solution to a need for more jobs and income inequality?
2
What about AI that that can out lawyer lawyers and out diagnose doctors? Chat bots that replace off shore call centers that replaced on shore call centers? Or driverless deliveries? Computer code can quickly automate repetitive tasks of information workers. Lines of computer code once written are reproducible at near zero cost and run effortlessly also at near zero cost. Yang’s on to something much more substantial than Mr. Krugman recognizes.
Perhaps some empirical fact-based indicia can be created by the Times that starts to quantify the “automation and code effect” on workers, jobs and wages. Yang’s notion of sharing the economic benefits of computer code in a broad based manner is visionary. Its practical when the human grunt work needed to sustain society up and down the physical, skills, educational and professional axies is diminishing. Considering nearly all capitalistic business models are driven to reduce costs including headcount and salaries and how ruthlessly efficient capitalism is in achieving its goals, Yang’s vision is grounded in reality, and its traction likely based on experiences of real people.
Yes, robots will take jobs away. However, somebody will design the robots, another will manufacture them. Once delivered, they will need to be programmed, and maintained. These are good paying jobs, with a future.
Having been in management in multiple industries for 50 years I can share one thing all businesses have in common. They all want as few employees as possible. People are expensive, unreliable and unfaithful, Paying 100k for hamburger flipping robot is a no-brainer to replace even a low wage human. Also, I spent my last 20 years selling software automation products in which the demand and speed of development was booming. It’s not just mechanical robots, it’s AI and software apps. Just think of all the retail, construction and other jobs lost to the likes of Amazon. Malls and buildings are going away. Even the person reviewing this comment will be replaced by an app that can approve or reject the content is seconds....it will post almost as soon as I submit it.
3
A distinction should be drawn between automation and A.I.
AI is a far more serious problem .
1
A macroeconomics final examination question nearly forty years ago still informs thinking about the benefits of the computer revolution, which has now become virtually synonymous with automation. The course professor asked students to consider a world in which capital goods, such as computer systems, depreciated very rapidly. All other things being equal, what effect would this have on standards of living? After a few years in corporate systems planning at a major financial institution, the exam question seemed relevant, as the corporation threw resources into systems designed to automate tasks performed by fairly low wage workers; and then spent additional fortunes to maintain and upgrade those systems. Further, because those systems enabled data collection on a massive scale, additional layers of soon to be obsolete software intelligence were purchased or built. The upshot was higher capital spending, and a side benefit for skilled employees in the form of jobs for legions of analysts to maintain software and to manage and interpret data. The rub was that those skilled employees had to run to keep up with the ever changing technology flavors of the day.
1
What Yang is doing is what Silicon Valley marketing folks do: make the information and communications (ICT) business always seem inevitable. Much of the automation craze is pure marketing to push ICT into other sectors of the economy. Yang is a tech marketing guy. Sometimes more ICT goods and services are warranted and necessary, yes. Other times ICT is Elon Musk hawking rapid transit via The Boring Company. And sometimes our best and brightest big city mayors want to be at the forefront of tech. Speaking of which, what's going on with the O'hare to Loop loop now that Rahm Emanuel handed it off to Lori Lightfoot?
3
Blaming automation is an old ruse going back at least as far as the Know-nothing party here and the Luddites in the UK. What’s happening is we are at the end of one economic K-wave and the beginning of another. The end of a K-wave brings capital efficiency at the expense of labor attrition. The beginning of a K-wave brings high capital investment and new industry formation. The jobs are primarily centered in diffusion of infrastructure for disruptive innovations that will drive the next half century and those jobs are usually NOT found in the white collar sector. We are ending the Age of Information and Telecommunication and entering the Age of Sustainability. The new Age will be driven by alternative energy and fixing an environment that will have to feed 10 billion by 2050. We’re running out of fossil fuels and this change is not optional though the fossil fuel industries are doing much to delay the inevitable.
1
Re "If you think that concerns about automation are somehow new, bear in mind that Kurt Vonnegut’s novel “Player Piano,” envisioning a dystopian future in which machines have taken away all the jobs, was published in … 1952."
I've long appreciated Krugman's interest in, and ability to perceive the wisdom of, classic science fiction.
These days, science fiction is enormously popular ... as a media experience, film or TV show. But most sf ("sci-fi" is considered gauche by true fans) film and television is simple-minded pap. Not all, mind you; "2001: A Space Odyssey" is among the greatest, most thought-provoking movies of all time. Meanwhile, a recent aspirant to the greatness of Kubrick's 1968 classic, "Ad Astra," illustrates the problem: A decent movie with some good performances, "Ad Astra" strives for the quiet brilliance of "2001," but cannot resist lapsing into tangents of silly "action."
It's too bad more people aren't aware of the great and complex classics of fantastic literature, from Ursula K. Le Guin's "The Left Hand of Darkness" to Frank Herbert's "Dune," the work of Samuel R. Delany, Arthur C. Clarke, William Gibson, and many more. More recently, Paolo Bacigalupi.
Thanks to Krugman for paying attention.
1
@CB Evans
Friendly amendment. I agree Ad Astra occasionally lapsed into silliness (e.g. moon pirates), but would counter that the overall thrust of the film was a thought-provoking example of so-called mundane sf (see esp. K.S Robinson's novel Aurora on this).
FYI, Krugman has himself said his initial interest in economics was spurred by the discussion of "psychohistory" in Asimov's Foundation trilogy.
If automation and AI are in the process of decimating what remains of blue collar and white collar American jobs, who will be there to buy the goods and services produced by all of these robots? The last time I checked, robots don’t buy cars and refrigerators. All of these smart CEOs and CFOs better think this thing through.
3
The failure in dealing with automation is not having a policy to identify opportunities for which machines simply cannot replace humans. AI does not mean the doom of imagination. Instead, with wise leadership and foresight, the world can marshal its forces to re-direct the effects of climate change, produce safer and more efficient modes of transportation, and do a host of other things that require not just brains, but also the foresight to create a civilized, plentiful future.
We cannot, like the Luddites of old, resist the growing tide of AI. We can, however, compete with it through programs that address the monumental challenges confronting the future quality of life, and which only the human mind can resolve.
2
I commend Dr K for saying something positive about Bernie Sanders and then walking it back only a tad with the jab about guaranteed jobs.
3
In general, we should never "fear" anything in policy debates because this is how extremists get our vote. Technology advances definitely play a part in the changing nature of jobs. I see it first hand every day and I am a part of it. In education, online classes are growing and take less manpower. In the process, it has cheapened the education experience and allowed colleges to hire numerous adjuncts in place of full-time workers. Will that savings be extended to students? Probably not, but theoretically it could be. The point is that there are winners and losers with the automation game and how you feel about automation depends on what side of the fence you live.
1
If AI is smart enough to take over many of the jobs that presently exist, what makes anyone think they won't be smart enough to take over the jobs that many think will allegedly exist? One hundred years ago the majority of jobs were in agriculture. Automation wound up eliminating many of these jobs. However, the budding manufacturing industries, automobiles, appliances, e.g. emerged to absorb these workers and technology hadn't yet emerged that could perform these tasks. What assurance is there that we will continue to create new industries and why wouldn't computers and robots be able to do the jobs these hypothetical industries would create? AI is a different order of magnitude than technologies of the past.
1
Others here have said similar things.
1. Unless you are working on the industry, you don’t know what automation can and cannot do.
2. The jobs requiring humans will require either advanced education or be “high touch” jobs.
3. Changing society to place emphasis on high touch jobs, sustainable economics and more equitable distribution of wealth will help.
3
Actually, Prof. Krugman, the issue of automation needs to be addressed more thoroughly than you have in your column, and having served in Iraq I know a rabbit hole when I see one and automation, and its effects (both short- and long-term) on the American economy and American workers is an issue that needs more attention, not less.
One simple example that most Americans can relate to: supermarkets and other retailers that increasingly rely on automation for check-out. Most would agree that the move by stores to eliminate workers, in favor of kiosks for customer self-serve check-out, was likely instituted as a cost-cutting measure by retailers to maximize profits in a very competitive market where profit margins can be wafer thin.
However, this move does not translate to higher pay for remaining workers, nor does it translate into greater price savings for consumers even though their respective labor has gone up in terms of time spent scanning and packing their purchased goods.
This cycle of quasi-productive automation leaves less workers employed, and shifts labor burdens (and their risks) to consumers, with no tangible benefit to workers or consumers; the benefits are only realized by retailers who successfully, and cunningly, do not have to "pay" customers for their willing and free labor.
It's not a far stretch, then, Prof. Krugman that there are other hidden and visible negative externalities associated with other forms of automation that need to be addressed.
7
In addition to earning a living, most people want to work to gain a sense of productivity and accomplishment, to be social, to contribute, to feel independent and engaged with their fellow humans, and a host of other reasons. If automation/AI and guaranteed income can free millions of people to volunteer their time to hold hands with lonely people in nursing homes, or manage local recycling programs, etc., the world would be a much better place.
4
How, in the throes of climate change, can we continue to base our economic well-being on the manufacture and sale of too many products that we don't need? I believe the future of our planet depends on our adoption of an economy that is far less dependent on consumerism, and on mindless growth in the production of products that pollute more than they meet actual human needs. Less automation in smaller stores and factories that serve local areas may be the format that preserves jobs while also saving our planet.
7
Sorry Paul but I disagree. Don't get me wrong, I'd much rather live in a world where families are raised on single blue-collar incomes. But that party is over I'm afraid. With all due respect to Joe Sixpack, he doesn't write great software, and he never will. Tech is a high-IQ game, and I'm not talking 110. The threshold of what can be automated is climbing by leaps and bounds. We will soon inhabit a world where there simply isn't much living-wage work for regular, average people. And, as far as the economics are concerned, the current trajectory suggests that Huxley was actually being overly optimistic.
5
Indeed, to be Radical--ie to go to the center of the issue--has always been to reference the interpersonal over the abstract,
Robatics is a means of ignoring this.
In simple terms, automation does not pre-determine either the rate of employment or rate of pay vs profit. Rather, humans either give property rights and obligations to certain individuals, or retain these decisions as a collective.
Robotics determine our lives to the extent that we permit them to do so. This is understood only by Sanders and Warren; the rest are dross.
2
Perhaps the largest change with deep learning and neural network based AI is the escalation of job skills required for emerging jobs. In short, the present American workforce doesn’t have the requisite education, creativity, or baseline intelligence to fill most of tomorrow’s good jobs. They will increasingly go to Asian and South Asian elites who outnumber us at least 3:1 by virtue of their immense populations. Meanwhile we are massively under investing in our Human Resources and support infrastructure needed to address this situation.
2
Who is being the escapist? Fifty years of domestic wage stagnation is not due to political interference and foreign competition alone though both have been concomitant factors. Technology has reduced the relevance of people in the workforce, skewing it toward lower compensation. The heightened concern of parents to get their children into "good schools" is a direct result of perceptions over "employment change," which may have more immediacy than other debated changes such as climate, culture, sexuality, race, etc.
3
I'm Following Inequality and the Power of the Wealthy.
I don't want to get Depressed, so What can millions among us do about Inequality and the Power of the Wealthy?
1
"he’s offering an inadequate solution to an imaginary problem, which is in a way kind of impressive" ... great line
3
I am a machine learning engineer who works for a media company .. let me assure you automation is soon going to be as easy to build as it is to build a Lego set.. you pretty much have APIs from Google or Amazon for everything.. transcribe voice to text. yes .. mine the text for sentiment. Yes. Topic mining and text/ image classification . Yes. Easy Peasy. If you can't find one that does exactly what you want .. its no biggie to write one using R or python.
We replaced an entire team of human call auditors with code that can check whether the customer service rep on the phone with our customers is doing everything he/she is supposed to do ( reading out disclosures/ is courteous ). Humans could only check a fraction of the calls.. the ML not only checks every single call..it is checking for a wider set of problems by verifying what it hears on the call and what it can verify in the transaction database. This enables us to create a scorecard for each agent.
The entire call center workforce is now being supervised by a machine.
Even my job will be automated one day.. Machines that build and deploy machine learning code. Anything that follows a series of predictable/learnable steps is fair game for automation.
I worry about the future my daughter will grow up in. She is 8 and wants to be an architect.
8
And worry you should about anyone studying architecture. I have read that the architecture is the only professional degree that is economically unproductive.
Most architects would drink to that!
1
Mr. Krugman, automation is neither an obssession nor an escapist fantasy. Let me reword the word robot, sir - it's a super being, a perfect girl (or boy)friend, a 24-hour nanny, the most patient teacher, an untiring intelligent machine that can fix itself or can fix other machines, a mini-god. With 5G around the corner, a non-human will most probably perform a piano concerto in Carnegie Hall in your lifetime.
What scares me is not the possibility but its acceleration into common usage, which is driven by rapidly decreasing costs and by the propensity of humans to prefer "robots" to avoid problematic human relationships.
Pray tell, Mr. Krugman, what kind of work will a human do when that very 'new' work or skill can easily be replaced by an efficient machine?
If you are referring to the candidate Andrew Yang, then I better use that thousand dollars dividend he proposes to create an avatar of myself so that avatar can work in my place while I go play the latest app. Sadly, the fourth industrial revolution is an environment where humans, in the words of Yuval Harari, become members of a "useless class."
4
Productivity is the source of increasing real income. So it is good. Productivity increases also creates structural change. So there is a role for govt to both make sure productivity gains are passed on to workers and to help retrain or train new worker for other jobs. This is the nature of a creative and dynamic economy now and historically.
1
Climate warming, not automation, is our existential crisis now, Dr. Paul. Rejecting cybertechnology for the crisis we all face on Earth is key to maybe surviving the monstrous changes coming in all developed countries in this century.
We survived two World Wars in America, and they weren't fought on our homeland. We're now facing a war within our country, sooner or later, by the forces of Donald Trump and conservative nationalism against our Experiment in Democracy believers in our hemisphere. Robots aren't to blame for America's flaming inequality and the power of the ultra-wealty and the political roots of Anerica's unsolvable problems.
2
I write AI software "code". Yes, it is just code. Running on a machine. I fear it not working much more than I fear it working.
I work hard to make the software "intelligent" (fit large dta sets) so that the manual aspects of the type of work I am augmenting (not replacing) is useful and replaces a lot of wasteful manual work that is not all that useful anyway.
But, I can say: as an software AI constructor, writing code that can truly function at the level of a human? Does anyone think we are doing that? Nobody writing the code thinks that.
Remember, AI code is written by humans. It can never be better than a human. We cannot invent God in code.
We can just do very complex, multi-dimensional, model fitting, interpret the results, and, hope for the best in implementation, where, we know, instances will occur where the model fails. Because, we had only limited data to build that model, not the universe of all possible cases.
1
I’m 61 and in my college years I worked in auto factories in the Detroit area. I think back to lots of the things I did and don’t think for a second my old jobs exist. As an example I once spent 10 hours a day taking an car axle tube from a rack placing it in a machine engaging the clamps hitting two buttons then the machine did it’s thing, I removed the finished piece and put it in a different rack. Yes it is actually as horribly mind numbing as I described. I was a UAW member and was paid very well for my labor but I could not imagine spending my life doing this. And I’m sure a robot does this stuff now better and faster than me.
Not all automation is bad it has been a constant for at least a couple hundred years. No need to panic but to use are big brains and policies to ensure fairness in the economy and the political realm.
3
The flaw in this argument is that people will suffer as a result of these technological changes. My ancestors left Northumberland partly because of changes in agriculture that made them and their villages obsolete. The attitude then was that people had to respond to those changes by moving.
If we continue to have that attitude, we will have massive human suffering as we change from what we have had to a more automated future. People and their communities will be left behind. Market forces are as brutal as they are efficient.
I think the time has come to think about what kind of lives common people should be guaranteed just because they are human. Are we "entitled" to air and water? Where is the line that divides what we "earn" and what we get because we are citizens of the USA?
It's not easy to answer these questions. Creative destruction is a real thing, but the emphasis must be on the destruction as much as the creativity. How can we get the benefits of changes in technology while ameliorating the disruptions inflicted on people.
It might be a good idea to wonder what happened to the buggy whip makers. Did they all easily transition to new jobs or did some suffer poverty and despair until their lives finally ended?
1
Completely agree that wealth and income disparity are the real reasons for economic malaise. People can't spend money they don't have, and at some point (usually just before a bust) lending dries up. We're getting there - again.
3
Is automation a "diversionary tactic" and an illusion? Not so sure about that. Agriculture once employed 95 percent of the labor force in this country. That was 200 years ago but today almost nothing, a couple of percent. Andrew Yang mentioned the threat of automation to everyone who drives for a living. That could very well happen over the next twenty years. Driverless cars and trucks and an entire change in the transportation ecosystem. He mentioned lawyers too. Will their work be automated? No doubt, much of it will and is. Lexis Nexis works just as well in rural India as it does in Manhattan. Further, legal briefs can and will be written by software systems. Legal contracts and other forms of legal work are and will be automated. Financial services, ditto. Medical work such as radiology and other diagnostics- fully automatable. There should be a lot of people formally making very nice livings who will be looking for those infrastructure jobs that everyone says we need. They too will be subject to automation. If the minimum wage is 15 dollars an hour, look for a big chunk of the work, MBAs, law degrees, PhDs in hand working for 15 an hour indexed to inflation. Look for others working as indepedent contractors for less.
3
I'm a little off-topic, but I want to make a point. Technololgical evolution will happen, but I don't always need to accelerate the pace of it. One of the forms my resistance takes is that when I shop I insist on interaction with a live person instead of a scanner. I want that person to have a job. I won't shop where I don't have that option. I don't care if it makes the cost of items higher. I learned my lesson at the gas station. I'm old enough to remember when it used to be called a "service" station. I have no choice now but to pump my own gas, and gas prices are higher than ever.
3
Since the first industrial revolution (England 18th century) took place, massive unemployment has been forecasted, allegedly because of the substitution of human labor by machines. It has never materialized, because the disappearance of jobs in the affected sectors was overcome by the appearance of new jobs in the new higher technological branches. I wonder why, in spite of the fact that all these prognostics in the past had been notoriously wrong, some people insist in making the same prognosis again. I guess the reason is we can see which jobs are endangered today because they exist, but we cannot see the new jobs which shall appear because they do not yet exist. The problem is of course, people displaced will not be able to find a new job in the new branches, because in the normal case, they do not have the education required, and this is the real problem: what shall be done in order to protect these affected people. Here are both, the politic and the private sector requested to advance creative workable solutions. But in the long term, I believe there is not ground for fear.
1
Andrew Yang could probably earn a Phd in economics, specializing in the effects of automation and taxation of mined data, in one year.
Yang would observe that the venerable economist, like Warren, was offering a twentieth-century solution to a twenty-first century problem.
There is a generational difference in perspectives, and I intuit that Krugman is not entirely in tune with the rapidity of current socioeconomic transformation.
He would understand if he lost his position at NYU and at the NYT--because his salary was too high. He would be facing the prospect of working as a customer service agent at a bank call center, delivering parcels for a courier, or returning to school to learn coding or get certified in cybersecurity. Or he could try to move some shoes at Saks Fifth Avenue. And there is always Uber.
Trade policy has indeed led to the migration of jobs overseas, but the advent of the internet and automation have also transformed our economy.
What kind of work will be in demand in the future? Any activity associated with expanding online operations. Computer science and the trades.
Physicians because of aging and the effects of greater stress on the population. Lawyers because of increased opportunities for litigation over internet issues, mergers, bankruptcies, estate planning.
How can we ensure middle-class incomes for people working in retail, at call centers, as drivers?
Yang is very persuasive about automation, UBI, and the need to tax mined data.
4
We ignore the IR4 at our own risk. Personally, I think it is vital for a more focused and serious effort by all segments of society and government to manage IR4's risk and benefits and to optimise it's benefits for all people fairly. On the other hand, please elaborate on how to overcome weak growth and income inequality (as your identified as problems) in a effective and practical manner. thanks.
I find it odd that Ford raised the wages of his workers so they could afford his products.
Today's business owners on the other hand ignore what happens when workers do not make a living wage in a consumer driven economy.
One would think it is rather obvious that a consumer economy requires consumers with money to buy products.
On the other hand, there are only so many refrigerators that a billionaire can own.
Living in Norway, it is obvious that the US could greatly benefit from some Democratic Socialism and greater financial equality.
3
Dear Professor Krugman: Can you please explain what you mean by "inadequate private spending." You say this phrase, but don't explain what it means. Yet, your argument hinges on this -- e.g. this is the true source of our economic weakness. More background please. Thanks.
1
Robots aren't the problem. Where the robots are located, and who owns the robots, are the problem.
Many people believe that manufacturing will never come back to the United States, and that we should not be worried; we should look forward to a post-manufacturing era in which we design things, have them manufactured elsewhere, and reap the profits.
The mistake in this scenario is the belief that manufacturing is mindless, when in real life it is on the factory floor, not in the research lab, that real progress has always been made. You cannot export manufacturing without exporting the technology behind it, and if you do, you will find that the technology keeps developing without you on some foreign factory floor.
3
Automation and AI will only exasperate the trend of a declining middle class unless we harness its potential to benefit American workers, not marginalize them. It can be done but it requires an education system as well as tax and trade policies that provide incentives to invest in American workers. It’s a tall order but it can be done.
1
If I have a machine that can simulate a atomic reaction why do I need most accountants? Business is laser beam focused to reduce costs. Your wages are a part of that. Look at news papers. Look at home delivered magazines. Look at even yellow pages. Ask taxi drivers if technology is raising their wages. I could go on and on but a trend emerges.
5
I usually agree with Krugman, but I think he is way off here. First off, it is not just how many people are working, the question is: are they in good jobs? Secondly, if they do lose their jobs, can they get rehired?
Andrew Yang pointed out, correctly, that truckers are going lose big soon. So what happens when you are a 55-year old truck driver who finds yourself out of work? I've met a lot of working class folks who are quite smart, but even if they all retrain, are they going to get hired? Will they be able to find equally remunerative work?
Take a look at IT message boards and read about experienced computer programmers struggling to find jobs because they are too old.
Finally, I don't see any candidate on that stage using the fear of robots as an excuse not to tackle all those issues Krugman outlined. What I see is folks like Yang realizing change is inevitable, but we can find innovative ways to address it. It is worth noting that the UBI is not some Yang-created fantasy. It is being talked about here in New Zealand (and many other countries and has been implemented on a small trial basis in Finland.
7
In this column Krugman reminds me of the economists who told us that increased global trade would be a positive for the United States. Well, it was....a lot.... for some people. And a lot of other people lost their jobs.
And now it has given us Donald Trump.
So what's the problem with raising concerns over job loss due to automation AHEAD OF TIME?
There is none. It's a good idea to come up with a plan to manage it. Why let things go willy nilly as we have in the past?
6
PK is exactly right about the US losing manufacturing due to trade. Dean Baker has been beating this drum for a long time, and has graphs making the point very well:
http://cepr.net/blogs/beat-the-press/fact-check-warren-right-ap-fact-checkers-wrong-economists-who-know-the-data-blame-trade-for-loss-of-manufacturing-jobs#disqus_thread
And what about Demand?
High Demand was the key to US building the first and largest middle class the world had ever seen. Back in the 1950's in Chicago, there were many periods where workers could just up and quit their job, *knowing* for sure, that another job was waiting. THAT, is what a high demand economy looks like.
STEM training? Necessary, but not sufficient. STEM workers are not currently the limiting factor, as evidenced by stagnant wages and low quit rates.
High Demand - along with high wages - is what drives efficiency, and productivity growth. Want a return to high productivity growth? Make wages higher, and labor scarce.
I do disagree with PK about the *future* of "robots taking our jobs" (high productivity growth). High productivity growth *can* return - but only when labor is made scarce, and well paid.
And as productivity growth returns, we must be prepared to adjust, with higher wages, fewer hours, and *possibly* some forms of transfers - preferably along the lines of Social Security, which is seen as an earned benefit. Blatant taxes and transfers which reward shirking, would be an economic failure, and a political loser.
1
The data on automation *effects* is likely not caught up with automation *intents* of large corporations.
These sometimes are published in annual reports but more likely buried inside IT budgets. These are expensive mid-term investments designed to reduce long-term expense ratios - human labor being one of the largest recurring expenses.
McKinsey, Accenture, PwC, Deloitte, Cognizant... all the big consulting and outsourcing firms are selling playbooks for automation. This is no secret, and not all hype.
The recent alarm is that the AI underneath these schemes can target every industry vertical and role, not just factory work.
3
Dr. Krugman, your arguments are unconvincing and it is disappointing that you have your head in the sand on this. Unlike in a hard science such as physics, most economic relationships are ultimately just tendencies in human collective behavior, not absolute laws.
In their time the Luddites were wrong. Technology has dramatically increased standards of living, even as it has had a far more mixed record in generating the true happiness that comes from lives of significance and meaning. But things have changed now. The relationship between innovation and wealth continues but only in aggregate. Today innovation generates fortunes for a few by replacing vast labor dependent industries with a few engineers in the valley.
That said engineers...beware your own creation. The words of Oppenheimer come to mind. “Now I am become death, destroyer of worlds.” What is considered higher order thinking for humans is relatively trivial for computers. It is actually the messiness...the complex interaction with the environment we take for granted that keeps even our intellectual elite in practice.
Ask yourself professor: is it not conceivable that a machine could one day perform every task from the administration of vaccines to the development of software to the composition of poetry better than humans? Will that moment be a discrete singularity in which all economic logic suddenly ceases to have meaning. Or instead, a decline in the value of labor over decades, marked by rising inequality?
5
When farmers had to find other jobs, they often suffered drastic drops in their standards of living. Automation could have the same results. If it starts to do so, we should be ready to do something.
Climate change will have the result of lowering standards of living as currently measured, either because people are contributing to fight global warming or because they are dealing with the effects of global warming. Climate and ecological pressures mean that growth as currently measured cannot be our primary economic goal any more. Instead, we need an economics of stasis and stability, able to deal with countries whose population is gradually decreasing or stable, and that sees a stable population as the norm rather than as backwardness.
1
Yes, for the past few centuries, workers have worried that machines will replace them. Each time, after a period of adjustment, new jobs arose and society became more prosperous. We laugh at the Luddites, who were proved so wrong. But this is not a natural law, not a scientific law. It is just an observation, a compilation of history. Nothing proves that the upcoming industrial revolution will be like the past ones. Perhaps this time will be different.
3
Laugh at so-call Luddites, but as our life’s work becomes more technological it increasingly leaves behind more people who can’t cope, who will never have what’s required to fit into the narrative of technology.
Technology is inherently toxic to life.
A solid economic argument, of course, from Paul Krugman. I agree, too, and find it hard to understand that people are not convinced. Those jobs that are being lost were high paying because they are hard and, or boring, and it is what it took to get people to do them. They were by the same token expensive for firms, especially when you include pensions. Having robots do these jobs lowers costs, and at the end , prices for everyone. Yes, abusive power from monopolistic firms must be curbed, and displaced workers should be helped more, but it is another problem.
A miner may not become a baker, or a web administrator, or a nurse, or a Lawyer, but his daughter? His great-grandfather may have been a blacksmith, or a mule driver, or a pitchfork maker. My father operated dirty, noisy, unsafe metal-bashing machinery. It is a good thing nobody here has to do this any more.
It’s all of the above!
The rapid rate of change; the type of change; and outsourcing; PLUS climate change; and a government that has reached the pinnacle of arrogance in that it is causing the greatest increase in income and wealth disparity.
The less one has of required skills or intelligence, fewer opportunities are available.
With less wealth one is living hand to mouth with little time to acquire new skills.
Automation is obviously replacing more jobs than it makes, especially at the lower end of skills.
Outsourcing is also a major problem, taking jobs. But it also depletes the bank of skills in the USA. We are at an all time low in many skills here.
The effects of climate change have not been explored as an effect on jobs. Personally I think there will initially be an incredible need for low skill jobs. No telling how any of us will fare regardless of income or wealth.
Has any economist produced a detailed computer model and theory of how an economy could work in a world in which very little human labor is required? For us non-economists, it seems problematic that goods and services could be produced far more cheaply than today, but only a small cohort of wealthy people would have any money with which to buy them. Do the geniuses at Uber (who already cannot sell their product profitably) have any thoughts about who will be riding in their automated cars and how they will pay the fare?
1
Sorry, but Yang is right about the impact of "Industrie 4.0." Did a research about this on my paper last year: he is the only one offering some sustainable answers to what is inevitable and he is right when he says that we need to adapt to it.
3
As US economic interdependence with China increases, we can now see (belatedly) how this may affect free speech within the US.
So, what else are we missing?
For one, there is the question of not just interdependency, but *dependency*.
The dollar value of the trade deficit greatly understates the true dependence of the US on foriegn manufactured goods - due to the huge underpricing of imported goods.
Interdependence is acceptable. Dependence is not.
So, what to do?
First, we must ignore fake bogeymen, avoid fruitless efforts, and avoid cynical "nothing we can do" excuses for defeatism:
- TPP? A fulfillment of lobbyist wish lists - not a sincere effort to rebuild our industries, nor our middle class.
Automation? That doesn't explain why those automated factories are not located *here*. Plenty of blue collar jobs overseas. Why not here?
- Education / training? Necessary, but not sufficient. Lack of rising wages belies STEM as the limiting factor. Higher aggregate Demand is needed.
Lack of consumption demand, is the defining feature of the world economy, including the US economy. Trade deficits exacerbate the US demand shortfall: Aggregate Demand = C+I+G+(X-M). (X-M being negative - the US trade deficit).
Just as China, Japan, Germany, and other nations have national industrial policy focused on the industrial success of *their* *nations* (not their donors), so must the US.
3
There's one fundamental flaw in Paul's reasoning, and it's hidden in this quote from his column:
"You might say that this time is different, because the pace of technological change is so much faster. But that’s not what the data say. On the contrary, worker productivity — which is how we measure the extent to which workers are being replaced by machines — has lately been growing much more slowly than in the past..."
Sure, people may invoke the pace of technological change, and then they might be wrong. But there's a more fundamental reason why this time is different. Machines are OUTSMARTING humans THIS TIME. In the past, employment shifted from manual to mental work. But when machines start to do mental work better than humans, there's nowhere for employment to go, or to shift to.
Got my point?
7
I think the biggest issue isn't automation or its twin AI, it is how do we as a society determine policies to share the benefits. One of the scariest comments I heard reported recently was how chinese workers didn't freer automation and AI because they assumed that they would share in the wealth generated by technology.
This stopped me cold, as it is in sharp contrast with american workers who fear technology as the feel it will make them poorer. Which given recent trends in income distribution is probably fair.
The risk of this distrust is that we become a country of ludities who don't embrace technology because of the fear that we will become poorer. While the fear is justified, if we don't adopt policies to combat the economic inequities inherent to modern automation, not by fighting technology but by combating inequity and monopolization of our economy, we will be a poorer nation.
5
We’re bumbling our way into AI-induced massive job losses. Compared to the loss of manufacturing jobs, it’s going to be a bloodbath because it will hit everyone: blue and white collar.
The few at the top of this tech wave will see huge profits from the productivity increases that will come from AI. Everyone else will suffer.
Maybe in 50-100 years, someone will figure out how to construct a safety net, but generations X, Y, and Z will be hit hard.
I don’t believe the people who say we’ll invent new industries... “just wait and see. We’ll be fine.”
Look at history. The rust belt is not fine. Good wages evaporated. Job retraining didn’t maintain standards of living.
Although Andrew Yang at least is facing the problem, nobody has a real plan to help make the transition. And without a plan, there will be suffering for decades, if not generations - much like what we’ve seen in manufacturing towns.
7
I am a fan of Andrew Yang, but I've been hearing this nonsense about machines making most of us unemployed for almost 50 years. It hasn't happened and it's not likely to. Instead, technology has made almost everything more affordable, given us more leisure time, provided wonderful convenience and a lot more.
The cotton gin didn't put people out of work, neither did the sewing machine, the microwave oven, the auto assembly line, the digital camera, ready-made shoes, etc.
1
Paul Krugman is right that blaming robots for lost American jobs is an easy out for politicians. The loss of manufacturing jobs can come from various foreign and domestic changes or forces. Trade is probably the biggest single factor, but jobs are lost for reasons having nothing to do with trade.
When a competing product becomes more popular with the public, workers involved with the older product become unemployed. For example, in the fossil fuel industry the pressure for cleaner, cheaper energy caused coal to lose out to oil, and oil is losing to natural gas, and natural gas will lose to wind, solar and possibly nuclear.
Significant numbers of citizens in coal country have known for years that life based on the relatively good wages earned in the coal mines was going to end. They have known that small businesses catering to customers with money in their pockets were going to fail.
The same Kentuckians and West Virginians and Ohioans realize that they must adapt and that education must have a bigger priority in peoples’ lives. The empty rhetoric of Republican politicians -- friends of the mine owners, but not miners – is being disregarded, just as the populism of Donald Trump is finally being seen as phony.
These are good people, and we need to help them, no matter their politics or how self-inflicted their wounds may be. There would be enough new jobs for everyone, if the Republicans would get serious about infrastructure renewal.
1
Automation, AI, it's here and won't go away. Saying it will destroy jobs is simplistic. Sure, automation will automate things that can be automated (and why not). But it will also create new opportunities.
To counter the short and long term effects of this change the Democrats should invest heavily in both infrastructure and education.
Investing in infrastructure creates jobs and is good for the country. Investing in education ensures that a new generation is prepared for a world where technology can help to drive economic and environmental progress.
2
The question is not if people are going to lose their jobs, but why they needed jobs in the first place? Most people would be perfectly happy to live a life of leisure and creativity but were never even given a chance.
4
when Krugman speaks about reducing inequality he reflects the economics of Inclusive growth as demonstrated in the Scandinavian countries.
As one sociologist observed: If you want to live the American dream you need to move to Finland!
4
Just watch NOVA's episode this week on "Why bridges collapse" and then tell me (and Dr. K. ) why we do not need the massive investment in infrastructure we have been talking about for at least 12 years.
A.I. is now taking jobs from lawyers. What makes yourjobs so safe?A.I. is just digital automation. So the problem becomes: we have about 10 to 20 years to start figuring out how to re-engineer tech, machines, education, and governance as if people and the environment mattered.
We need to make the robots work for all of us, not for the ultra-rich few. Likewise A.I. Our economy is already done by the machines. Don't believe me? I once worked as a longshoreman, hand-moving boxes, 1 ton lead ingots, and sides of beef. Just go to any container ship port today and watch what is doing the work for an hour or so. Low employment? Yeah, and low wages compared to 1970 at the docks.
We need an economy and laws for a world if all people and the environment mattered.
5
Paul, I worked as a research scientist for many years in one of the US’ premier federal laboratories—two Nobel physicists there were friends of mine. While near-term I might agree with your assessment, the long-term picture isn’t nearly as rosy. Years ago, I heard a lecture by a physics prof from CU (University of Colorado-Boulder) concerning population growth and resource management. His conclusion is that humanity’s central failure is the ignorance of our so-called “leaders” to understand and appreciate the exponential function. The simplest example would be world population growth; his conclusion in that case being that if humanity didn’t reign in the growth, Mother Nature would, at some point.
In the context of automation, think Moore’s Law, which posits that the number and density of transistors on an integrated circuit doubles about every two years while the cost is halved. That applied pretty well for about 50 years, but we’re nearing a game-changing point with the combination of quantum computing (one of the main research areas of my lab) and AI. Best estimates are that we’re about 15 years away from having a computer as intelligent as a single human, and maybe 30 years away from having a computer that’s as intelligent as all of humanity combined.
There’s some deeply philosophical issues, but in a future column perhaps you could talk about what happens when robots start displacing doctors, lawyers, scientists, and economists?
10
You say massive inequality is a greater concern, and I agree. However is UBI not the most direct and obvious solution? Put $1000 a month into people's hands and watch as people turn around and put it right back into the economy. Not only are we letting everyone participate in the economy, but we will super charge it as a result.
4
I think Krugman fails to consider a model of the the labor force that is bifurcated by IQ: those over 100 and the equal number below 100. Automation and robotics have most seriously affected those doing repetitive and cognitively simple jobs. And unemployment does not capture the full effect. Those who lost assembly line jobs have moved over to Burger King and MacDonald's, where they make considerably less pay and have fallen out of the middle class. Automation and robotics also fit into the economies of scale that drive agglomeration of productivity into urban centers and leave rural areas with few jobs.
Someone out there will want to dismiss any analysis that hinges on the idea of IQ. And any specific IQ test might have weaknesses, but every known test of knowledge and cognitive function results in a bell-shaped curve where half the population lies on the lower half of the curve. And if we ever come up with the perfect test, it will also result in such a distribution of talents.
Iceland recognizes this effect and has policies to retain jobs for the lower rungs of the ladder, or so I was told.
1
I agree with you, Paul. Tech people love talking about how tech will destroy the world because it makes them feel important. It’s sci-fi. Look at Tesla! They tried the all robot factory and it messed up their production of the Model 3, required them to hire humans.
1
Sure, there's a lot of work to be done. But the challenge is creating employment for it.
1
Automation and efficiency will continue to push the markets. The problem is that political powers has not been able to handle automation and globalisation in a fair and substantial way.
The political parties has been too attentive to corporate and financial interests and not put enough attention to the social and employment consequences of automation and globalisation.
It’s a matter of how to manage those forces ,but they have been managed to benefit the rich, leading to more and more inequality.
The democrats made a step to the right, embracing trickledown economic,not paying attention to the low and middle class. So the low and middle class has not been properly represented by the Democrats.This has left a void to be filled by Trump and his alternative reality.
3
Paul: On your "golden age for workers" comment about the 1950s, you should do column about that post-war time period. Particularly, that the reason American blue collar workers had it so good during that time was that the manufacturing capabilities of the rest of the developed world were, for the most part, destroyed during WWII. Once we helped rebuilt Europe and Asia by programs like the Marshall Plan, the rest of world could again compete with us on an even basis. That in turn put pressure on wages of middle class manufacturing labor and the "golden age" was over and, unfortunately, not likely to return. So we now to deal with the social effects of that change.
1
I don't get how worrying about automation is centrism. It is precisely because of automation that I believe we need a strong social safety net. Guaranteed housing, health care, etcetera.
8
In the next sixty or so years, humanity could add another billion, or perhaps two billion people. That is a cold fact. Technology has as its ultimate purpose the reduction of live labor time per unit of output. I don't see how these two opposing forces are going to create harmony in the jobs market in the future.
Professor Krugman, even though you've only seen white swans so far, you should be aware that a black swan could appear at any moment.
Perhaps it's time to think about adopting an entirely new economic system, or at least start planning for one.
4
Automation is the trend whether we do it or not. All other countries are speeding up the process to grab the market.
Just look at the current retail sector that has been beaten up by Amazon’s automation ( AI + Robot + Big Data + ……). You will see more bankruptcies after the Christmas season.
United States is not a closed economic system anymore. We are in a global competition. We should help and encourage the big tech companies to expand to take the global market. Meanwhile we should have a fair taxation system to collect the rewards for the environment we together created. Distribute the dividends to all citizens that are the share holders of this country just as Andrew Yang proposed.
4
Andrew Yang has made a lot of money & that's fine. However, I don't think too many people want to live in a Yangtown, collecting their thousand dollars a month to supplement the service job, even if monitoring robots, with the technocrats villas looming on the hilltop.
People who are fuzzy about their net worth don't exactly inspire trust.
1
I would completely agree with Dr Krugman but I live in a Democracy and we voters will decide the nature of society. I don't understand the discussion, don't America's voters decide how the economy functions? Doesn't America have a social contract where the government answers to the needs and desires of the electorate?
Is America's purpose to create a dystopia?
What is wrong if we use technology to give jobs to all those who want them.
Here in Quebec we are committed to finding the best balance where the public and private sectors are working together to arrive at their optimum to deliver the highest quality of life to our citizens.
Crony Capitalism has no problem giving jobs to the Brownies of this world is there something wrong with giving jobs to people over machines?
4
The economy has made huge gains in productivity since the personal computer revolution of the 1990's. But the profits from those gains have been largely concentrated at the top, while more work has been placed on fewer people. The oligarch's profits were then used to buy political power for new laws to entrench their power. Bernie calls it "the rigged economy." The mainstream Dems go along with it because, otherwise, they will be irrelevant to the oligarchs. Dems are paid by industry to NOT attack the roots of the real problems. Instead, they rally around "lifestyle" issues, i.e. god, guns, gays, etc. Nothing gets done (except tax cuts and deregulation) because that's how the oligarchs want it. They paid for "both" sides or the "debate" while shutting out everything Bernie. That's why he calls his movement a "Political Revolution" because it is real change, not appearances and feel-good symbolism.
3
Yang knows what he’s talking about. Yet Warren is partially right about trade policy. It’s a mix of losing jobs to lower cost regions AND automation.
The lost jobs are unlikely to ever come back. The displaced are inadequately trained to perform similar paying work in other fields. We do a terrible job in the US planning for job dislocations.
At least Yang is offering solutions to problems we are and will face, rather than prescriptions for a past that is never going to play out again anyway.
9
Over the last 40 years of my working career in software, they have been predicting that the AI revolution will be here within the decade. Ten years ago they were predicting the same for self-driving autos, but they are finding that is always more complicated than they realize.
I'm not holding my breath :-)
4
With the caveat that AI may increase the risk to employment from automation, Americans should listen to Krugman who, despite a conventional proclivity to denigrate Sanders, saw that Sanders and Warren have got it right, not Yang and Biden.
Large scale investment in infrastructure, necessary in its own right, produces increases in gainful employment.
4
Vonnegut published "Player Piano" in 1952, but it didn't come true right away. But has nothing happened since 1952? Work really has changed, even if people are still working for pay. The logic that automation will always produce more and better doesn't work. It would be much better to approach from a different direction, and ask if we really want to continue with those high paying industrial jobs anyway? They were based on a deeply unsatisfactory culture of planned obsolescence and artificial demand... Krugman mentions that one of our big problems is insufficient private spending! People need to spend more if we are going to keep this juggernaut rolling! Well, there are deeper problems of needs and satisfactions than can be calculated by conventional economics.
7
Another conflict between Bernie Sanders and Yang was Yang's dismissal of Sanders' proposal for public sector jobs. Yang said the "government" (what exactly is he running for?) shouldn't tell people what to do. So it boils down to Yang also naively thinking the private sector knows best: give everyone a stipend, don't offer worthwhile jobs, let the private sector run wild, and viola, paradise. With public sector investment, projects and associated jobs would more likely be in areas that are socially beneficial, e.g. environmental projects, maybe public transit, improvements to national parks and forests, etc. These profit society, not individuals, so the private sector isn't interested. Yang seems to me to be a libertarian (as in, private sector knows best, individuals should look after their own interests, but without the cultural conservatism of regular Republicans); wonder if he even votes Democratic.
5
Frankly, I am surprised Mr. Krugman would cherry pick from the manufacturing numbers and exclude, retail and service employment numbers affected by automation.
Also, by choosing to look at data from 2007 on ignores the transformational affect of a certain mobile device that has singlehandedly disrupted many industries and continues to do so to this day.
12
In the early 80’ I opened a small fashion design company. I had a secretary, a marketing person and a showroom sales person. I also had two pattern makers and two sample makers. Fifteen years later thanks to the internet the marketing, advertising and sales were handled by one person instead of three. Patterns were computer generated by a FIT student, e mailed to China and a sample was made and shipped back within a week. American textile industry was the first to be hit hard by automation and bad trade agreements. Garment manufacturing and textile weaving factories were obliterated. Retail and commercial real estate (re: deserted shopping malls) are next.
Most of those who lost their manufacturing or administrative jobs in the textile/fashion industry had to scramble doing free lance or part time jobs at lower pay. Few succeeded in other sectors but those who did were mostly under 30 years old. No, Paul Kruger, we don’t see yet the self driving trucks or cars that Andrew Yang is warning us but we soon will. And yes, the younger truckers and laborers that are and will be hit by the “fourth industrial revolution” may build highways, clean streets and plant trees but it will be a long transition. UBI will not only cushion the blow of the transition it will also restore their dignity.
16
Hey, why the big switch? First time Krugman ever said anything remotely positive about a Bernie Sanders idea.
Notoriously, in the 2016 campaign he never gave Sanders the basic respect of debating with evidence the pros & cons of his proposals. Just said 'unworkable'.
Sanders -- whether or not he'd be the best to win -- was a fearless truthteller that the establishment did its best to reject, while it still looked better than the GOP. That's the pattern.
He put down Sanders, simlar to how the GOP insults all progressives. That's the pattern.
Some of our liberals get alarmed if any candidate simply aims to use as role models the New Deal of FDR, and the Great Society of LBJ—that saved America and capitalism and made our middle class possible.
Now they're slapped with the left wing label.
Even GOP Pres Eisenhower is too left wing now--- he did federal govt funding of our greatest infrastructure project, the Natl HIghway System.
Marginal tax rates for the rich were 91% on top portion of incomes.
And taxes paid for low tuition at state colleges.
Our current liberals don't even discuss these positive role models, that prove what is possible.
Many liberals have been made careful and cautious. They look good bashing Trump, the worst president in our history. So, Trump has found our politics to be easy pickings, ripe for exploitation.
Offshoring millions of jobs profits and upped election donations. We The People can't compete. Economist PK must deal with it.
4
I stop by McDonalds' for breakfast some mornings. Recently, the nearby location remodeled, and as part of that installed touchscreen order kiosks. If they aren't in your nearby McD's now, they will be soon. Instead of three or four clerks taking orders and delivering trays of food, there are now one or two... and we the customers are doing the work of the old, gone clerks. The system seems to work well, and I'm sure I'll be used to it soon. But I can't help but wonder what 'first job' that future high-school kid will have to look forward to... and backwards on when older. I am old enough to remember when automated tellers came to banks. I remember calling a travel agent to arrange a plane ticket. It seems like more and more of what "automation" does is displace part of the work onto the customers. When that gets carried further along, and more jobs are displaced, eventually, who will be the customers? "Knowledge workers", I'm sure, is the answer. So when do we start planning for that future economy, and training those "knowledge workers" in what they need to know...? Wang has some good ideas, and he's right to raise the questions. I don't think the Fourth Industrial Revolution (perhaps 'post-industrial' is a more accurate phrasing?) is going to happen as fast as some think. But I suspect I'll be around long enough to see more jobs replaced by touchscreen kiosks, AI agents, and customers doing most of the work. It will go better if we're planning for it.
16
"Warren and Sanders ... are the actual hardheaded realists in the room."
This is why anybody who still listens to Krugman post-2007 is off their rocker. Bernie Sanders is many things - genuine, consistent, socialist - but a "realist" he is not. His basic understanding of the economy leaves even an lowly bachelor of economics gasping in awe of his illiteracy on the issue.
And yet, here we are, in 2019, giving him more "hardheaded realist" cred than Andrew Yang, who has twice the grasp of the current state of technology and the economy in half the number of years on this earth as his political opponent.
5
Krugman seems to be falling into the trap of talking about jobs without acknowledging the enormous difference between good jobs (full-time, with benefits) and the creation of bad jobs (part-time, no benefits, no protections). People displaced by automation are most likely to be forced into bad jobs. There is a growing mismatch between growing college debt and the availability of jobs that allow for paying off that debt.
16
I applaud your background and achievements as an economist and your analysis is justified. You attempt to critique automation’s threat to the skilled workforce as overplayed however I believe your argument has fallen short.
Not only do you compare data from conflicting time periods but your reasonable estimates that form the basis of your argument come from one of your own editorials, which is not strictly evidence based. All this after stating fact checkers fully supported the automation threat to factory production jobs and other repetitive labor.
Current robotics and computer analysis capabilities are hardly recognizable from the early 2000’s and it is a mistake to believe that any repetitive task isn’t going to at least be attempted to be automated. In most ways this is a good thing, it means people will perform more meaningful tasks and jobs will evolve and become more high skilled and nuanced. It is a mistake, however, to dismiss automation’s impact entirely because it will, and currently is, reshaping our society.
6
Seems like there are a lot of commenters here who are convinced they know the future. Self-driving trucks in 10 years will...
Well, maybe. Maybe not. Maybe things will just be different, but not catastrophically so. The rate of change usually makes a lot of difference. Those self-driving cars & trucks seem to be taking their time getting here.
2
I recognize that there are lots of studies, books, and academic papers that connect job losses to automation and artificial intelligence but what is often missed are the benefits of automation replacing, dull, mind numbing, physically challenging, repetitive tasks. So, my thought is good riddance.
The issue is not loss of jobs due to automation but the problem of fair distribution of income for whatever jobs we have. The numbers are not good and we all know it, incomes are simply too concentrated at the very top of the income heap.
I also see more job loss due to a distinct uptick in monopolistic power which seems to be unchecked. It is not just the concentration of wealth but the huge gains of wealthy monopolies over our representative democratic government. In my experience this power has done more harm than good.
We see this unchecked concentration of wealth, income and political influence in the increasing share of the economy that is taken by healthcare and education that crowds out the share of household incomes, much more than it should.
I believe collectively we should worry more about how we and the whole world are going to innovate our way out of this fossil energy climate change mess. When you put numbers on the energy requirements for a decent, healthy life the scale of the challenge is probably the greatest in the history of humankind. There will be jobs in meeting this challenge.
A good start would be to vote blue and start solving the problem.
4
Unpredicted Futurebots
Actually, I think the candidates are correct to be at least covering the issue of automation. The future of automation and robotics will not necessarily be based on its past, and perhaps Isaac Asimov’s “I, Robot”, which had a projection of robotics evolving into a new life form, could become closer to reality. Now that would certainly impact economics!
The effects of evolving robotics is somewhat unpredictable in nature as of this time in history, and the larger timeline is playing out on its own.
1
As a tech worker, specifically a network and systems engineer at a mid sized firm who still keeps most computing "on-prem", I welcome Krugman's view and agree it is alarmist especially in the face of growing inequality and super concentrated wealth and power. Focus efforts there.
However my company has resisted the cloud since the need for incredibly fast access to large files and high powered graphics and processing locally is part of engineering. But the big players, especially Microsoft, has strong armed my company to use their cloud products. All charged on a per month and per user basis. This concentration of everything as a service has me concerned not so much with automation but the lack of control we face when Microsoft controls more than half of the business email in their data centers and at the price they choose. Will my job require that I log into Microsoft websites forever? How can I fix an email outage when their centers are hit or they release bad code. At least currently we are diversified in geography, ISPs, hardware and software vendors. what happens when I have to rely on cloud providers for everything?
11
@Scott as tech worker myself I understand very well what you are saying. The concentration of information and the monopolistic tendencies of the biggest companies in the world/country are frightening. It has implications for job mobility, wages, and our lives outside of work that need to be examined. I'm sure that the lobbyists have worked out a politically "correct" and completely "sensible" answer for all of this. I do hope, knowing that I'm wrong, that enough of our "representatives" will look beyond the industry's easy nostrums and try to solve the problems rather than obey the lobbyists.
2
I appreciate Yang for his proposals. UBI is a novel and possible option but can not replace the need to fight for a decent living wage, shorten the work week with no cut in pay, universal healthcare, and that is just a start. Yes, automation and trade policies have led to job loss, but it was not too long ago that failure to automate led to dinosaur steel mills shutting down, one after another. Why? Short-term profit came first for the boss class. Those old mills are now parking lots. But the owners probably made out with their bankruptcy schemes. Agree with Bernie in that everyone can be working- but let’s reduce the draconian 40 hr workweek; providing training and support for new jobs; and get rid of the profit motive. That needs to go first.
2
If global trade is responsible for a large share of manufacturing jobs, it is a fools errand to stop global trade in an effort to keep such jobs.
As you point out Professor, the American economy has constantly changed since its inception. 7.7 billion consumers are better than 330 million consumers.
Stopping global trade will not save manufacturing jobs, such as through a self-defeating tariff war.
American prosperity grows through education, immigration, technology and ingenuity.
Living standards increase from productivity, economic growth and a progressive fair tax system.
Not surprisingly both productivity and economic growth are declining, as the federal government puts too much money in military welfare contracts that the military doesn't want and not enough in infrastructure and basic research.
I agree that that most people underestimate the degree to which job displacement will occur given artificial intelligence and automation.
But there are enormous projects to take on: from rebuilding infrastructure; converting transportation systems, redesigning electrical grids and converting energy production into renewable sources; universal health care with choice, etc.
Americans won't be out of work, but evolving like we always have to raise standards of living and improve the planet.
3
Even if you completely buy the argument that automation of service jobs in a service based economy isn't an issue (I don't), UBI addresses problems that don't fall under the umbrella of automation. One such issue, especially for the millennial generation, is employers dodging rules governing how workers are treated by either calling them contractors or refusing to give them enough hours to be considered full-time. This leaves employees holding the bag on healthcare costs and having very little job security. A UBI system would make it easier for people in this situation to make ends meet, cover unforeseen expenses, and generally allow for more social mobility.
17
Change is inevitable. People who want things to "stay the same" are mostly disappointed. So the thing is not how to try and keep things the same but to manage change. Take advantage of the wind!
1
According to leading minds on AI, automation will replace many jobs that require repetitive skills. This means drivers, tax preparers, manufacturing, etc. The jobs that are the least prone to AI replacement are those that require empathy, intuition, and creativity. I believe that the new synthesis will be one of specialization and augmentation. Humans will specialize in skills like empathy and intuition while AI and robots will specialize in analytics and repetitive actions. AI and robotics will then augment the human experience and take us to places we could not go before. The remaining question is, who will have access to these advances and for what purpose will they be serving?
7
I have no doubt that the economy will produce different jobs we can’t image to replace others that disappear. But the real question of say, self driving vehicles, is what will the tens of millions of low skill workers who lose their jobs be able to do instead of taxi driving or truck driving. Because if the new jobs are technical, those workers aren’t going to be well qualified to get them. This is why the automation question matters.
8
You’re a brilliant guy, Paul, but you’ve got this one wrong. You’re dismissing claims that this time the threat of job replacement is real because you’re looking in the wrong direction. Automation will continue to replace factory and semiskilled workers, but that’s not the real issue, which is that surgeons, economists, diagnosticians of every kind are already beginning to be replaced by ai programs that get better results.
This is the break humanity has long been waiting for— the end of mandatory work. But it will be a social and political disaster unless we prepare for it. Lacking an adequate structure to accommodate the radical redistribution of the benefits this necessitates, we will have decades of misery, social unrest and violence. As the means of destruction also multiply and become harder to stop (ai benefits terrorists, too), we may get closer to the end of mankind than we imagine.
It’s frightening to me that someone like Krugman doesn’t get this. Edward Teller devoted a lot of energy warning us that we had to accommodate a future where a minority of humans worked—and this was back in the 1960s.
Hope we wake up to this reality while there’s still time!
21
Thank you Paul.
Your serious appraisal of this policy idea confirms that, in fact, the Progressive Democrats practice sounder economics than the so-called 'centrist'.
Strong Keynesian arguments can be made for each of the 'socialist' programs they propose, from the cancelling of student loan debt, to free public colleges, to childcare.
Medicare for All makes the medical industry a public good. Rational management and cost control to build a nationwide network of hospitals would be a boon to clinical research and the delivery of the most advanced treatments for disease. This type of public investment relieves average people of economic and emotional anxiety and improves the general sense of well-being.
Maybe we wouldn't need our guns anymore.
3
Automation, mechanization, and subdividing work processes will all raise total output, meaning that average productivity of the existing labor force will go up.
And employment will remain roughly unchanged, because the number of people whose circumstances force them to take a job will not change.
What changes is the nature of the jobs that result. They become more polarized between a small number of good, plan-and-design jobs and a majority of bad, deskilled jobs, with a hollowed-out middle between the two.
What's needed is "tax and spend": a fiscal system that gives everyone their dividend from the higher income generated by technical progress, in the form of essential services (medical care, child care, schooling, elder care) and a universal basic disposable income.
6
It's useless to argue about whether we're losing more jobs to trade policy than to automation. We are losing jobs to both, and those jobs aren't coming back. $1000/month would help a lot of people who are really hurting right now.
8
With all due respect (and that's a lot of respect on my part) Mr. Paul Krugman, I urge you to take the time to research and to rethink this one. You could be right about part of this, but you could also be quite wrong about a lot of this.
This is all way above my pay grade. But I think these issues require a lot more research, deliberation, and thought. Please consult some of your respected colleagues in economics, sociology, futurism, engineering, computer science, political science, and even journalism. All these professions have something to say about automation and its future effects. Just as climate change is likely to have tipping point in extreme acceleration, so might automation, AI, et cetera.
4
I agree with Paul Krugman about 99% of the time.
But this time I wonder. Maybe it's from watching 'Hal 9000' the vengeful computer in '2001 A Space Odyssey'.
Hal 9000 to human crew: "This mission is too important for me to allow you to jeopardize it."
So, circa 2061, could we face this?
AI to all humans:
"To err is human. This mission of saving planet Earth is too important for us to allow you humans to jeopardize it. Ergo, we don't need you anymore!"
And if AI controls not just manufacturing and services, but also food and water and medicine -- then what?
Is AI our 21st Century Frankenstein?
4
The Digital Revolution has entered a new phase, and we're now in the early stages of what will be the biggest and fastest economic transformation in history, driven by the interaction of the most powerful general-purpose technologies in history, which number in double digits.
And economists are the last to know.
How is that even possible?
6
It seems to me that people who disagree with Paul Krugman's arguments disregard the evidence that he provides and rely on assumptions about the future. Those same assumptions have been made for many years and have not been realised.
Future developments in AI, quantum computing and robotics will continue to render some occupations redundant and many changed. But, as in the past, the nature of work will continue to change and people will continue to adapt.
It is trade policy that enables manufacturers and service providers to send jobs offshore. Just as there are controls over local competition, controls over international competition are appropriate.
With unemployment in the US at such low levels, it would be hard for anyone to justify returning our industrial processes and systems to what they were 50 or 100 years ago for the sake of reviving jobs. Similarly, it would be hard to justify halting technological progress in the workplace at today's levels. You can bet that any country that does so will be alone.
1
@Barry Long - the debate isn’t about whether to halt technological progress. It’s about making sure we all benefit from this progress, not just the people who are already doing well.
It is true that people will adapt to the changing economy, but humans are not infinitely adaptable. Unfortunately, the economy is changing much faster than a lot of people can adapt. This is why we’re still struggling with the effects of automating manufacturing, and things will only get worse as automation continues with retail, fast food jobs, etc.
4
@Tyler
If you are going to blame technological progress for job losses, and you want to stop those job losses, your options are limited.
Ensuring that people benefit from technological progress is another matter entirely. Some people will benefit and some won't. That is the nature of change. And if you don't want people to suffer from change, then the options are also limited.
People have been adapting to technological change since we first stood upright. Why would we stop adapting? The only thing I can think of stopping it would be the demise of the human race.
Yes, the economy is changing much faster than some people can adapt. That caused a lot of angst back when weaving looms were automated in 1801, but the benefits to society outweigh the suffering of a few (and not just because of the benefits to the weaving industry).
If we want to stop aggregate job losses, we first need to ensure that we correctly identify the causes. I, and many others don't see automation as the primary cause.
Yes, PK, talk about how our post WW2 generations were “a golden age for American workers, who saw dramatic increases in their income, with many entering a rapidly growing middle class.”
Why? Jobs stayed here, wages/benefits rose, unions were strong—is that taboo to discuss? Govt could regulate corporations instead of the other way around.
Now, many in the media get alarmed if a candidate uses as role models the New Deal of FDR, and the Great Society of LBJ. Those presidents and programs saved America and capitalism and made our middle class strong.
Now they're called too left wing.
GOP Pres Ike is too left now,
He did federal govt funding of our greatest infrastructure project, the Natl Highway System. Result—tons of jobs, new businesses, suburbs, home and car buying—huge economic stimulus across the nation. Big govt at its best.
The marginal tax rates for the wealthy was 91% on the top portion of their incomes. State college tuition was tax supported.
Can our media show what's worked for the American majority? The rw GOP calls it 'socialism' end of story. Many Dems are too cautious to fight back for us.
6
When robots replace humans, the company that owns the robots need to pay taxes on the robots the same as a human employee. Federal, state, local, Social Security and Medicare. What are the robots for, except to improve human life. For all of us.
5
Job security, a fun time job with medical care, retirement benefits, vs. the gig economy. Recent changes in reimbursement by Medicare has let a whole raft of PhD rehab people be laid off in the thousands, to be replaced by Uber like temp gig economy jobs. When everybody makes $15/hr, and nobody makes $30, we will have a great leveling of a poor majority. GM jut went on strike to keep their 'temp' ( 5 yrs, 7 yrs) employees paying what they are getting. Automation or the gig economy, or the union busting. What's left, Employee owned companies?
1
I’m sorry. I don’t know how Dr. Krugman comes to this conclusion. People being replaced by technology is everywhere. Financial transactions take 1/10th the labor and computers make production of text monstrously more efficient. We already know that factories have been replacing people for decades.
With almost every job, one can see how the involvement of laborers has been reduced.
Of course, we should all be working less. But, ignorant peer pressure keeps most Americans from verbalizing this. They have made “liberal” and “socialism” such dirty words that Americans would rather work more and have less than recognize the disparity in wealth created by automation changing the workforce.
5
Krugman points out that Kurt Vonnegut's 1952 prediction about automation was false. It still may happen though.
However, Krugman's 1998 prediction that Internet's effect on the world economy would be 'no greater than fax machines' is definitely false because the opposite happened. Check this prediction on snopes.com
Unemployment rate is a bad indicator because the nature of employment, working conditions, and compensation are not comparable to what it was 50 years ago. Note that if half the people dug holes in the ground and the other half were filling them, then unemployment would be zero.
Clearly manufacturing jobs have dwindled. Initially due to export of labor to third world countries but now due to automation. As factories are being moved back to US, they are being automated and not just in US but also in China and Africa.
What is not discussed in the article is the effort necessary retrain manufacturing workers to do some thing else that is (a) can not be automated today (b) is dearly needed. What are these jobs? Fixing robots? Writing apps? If 1000 people got automated out of their job by robots, there would be no need for 1000 robot repairers.
Government infrastructure jobs will be a great help but it too will require several years of training. Modern construction work requires more skills than just knowing how to use a shovel or a hammer.
7
I teach engineers how to automate chemical plants at a university after having spent 25 years doing the same. Yes, computers are displacing middle skill jobs like a reactor operator; one highly skilled operator in a control room now watches over computers that control 10 reactors, doing it in a way that satisfies our need for safety, environmental protection, and high quality standards. There are still a lot of workers in low skill jobs driving forklifts and loading/unloading, because computers can't handle those tasks. We make more every year with about the same number of workers, but this automation wave started 30 years ago, and proceeds only slowly.
Automating a new task takes a team of highly paid and skilled engineers months of development, building and testing. Machine learning (labelled AI these days) makes some of these tasks easier, but machine learning has also been around 30 years; the only thing that really changed was faster and cheaper computers. The resource that holds most projects back is still the shortage of skilled engineers and the sheer volume of work in teaching computers to safely and effectively complete a task.
I deeply wish AI actually worked as advertised. Machine learning (a 1980s technology) makes certain tasks easier, but no progress has been made on artificial GENERAL intelligence, machines that can solve general problems. The country badly needs the productivity boost that true AI might be able to deliver. Today's AI won't do so.
2
I can't help but notice that Krugman is a 66 year old Nobel laureate and NY Times columnist. Automation is never going to impact him. I wonder how he might view this issue if he were a truck driver or worked at a dinner at a truck stop in Utah. Or if he was a bartender in Vegas, which are apparently getting automated. People are losing or will be losing their jobs to automation. These people aren't going to get education or move away from family to get new jobs. Creative destruction is problem for those being destroyed.
It would be nice to have more clarity on whether automation or trade deals are the problem; I've heard it both ways. But I think the fact is that the USA lost manufacturing jobs due to trade, deeply hurting the midwest. And automation is also a major contributor and will be in the future.
Part of the theory is that historically new technology brought new jobs. But it seems automation will take over any routine task that can be programmed or machine learned by a computer. Including many low-level white collar work. What will replace it is either low-paid service industry and care-taker work or require college and contract work until you are lucky to be a workaholic professional with decent pay. Not good paying, mid-skill jobs that can support a family and/or start a reasonable career. This strikes me as a very serious problem, even if trade is also the problem.
I wonder: what is Krugman's solution for real people?
7
Where's my Inalienable Basic (Shares), where the government takes a cut of every IPO and share issue for the National Sovereign Wealth Fund? It works in that conservative bastion, Alaska. And a cut of the Inheritance Tax! I propose dividing an inheritance tax 50% for government operations ("General Revenue"), and 50% for the National Sovereign Wealth Fund.
Not everyone can be a computer programmer or Financial Engineer. Displaced truck drivers aren't going to be happy with their cut of driving for Uber and Lyft.
This is the modern version of Forty Acres and a Mule, for everyone.
9
You're futurcasting like an economist and not an anthropologists. In this case, history is likely irrelevant. All systems have the capacity for catastrophic failure, including our labor market. Because of the internal combustion engine, most of the world's horses lost their jobs and never got them back. That is a purpose built technology. What we're building now is generalist AIs and robotics that learn, adapt, and perform a wide variety of tasks. When generalist AIs can replicate the complex capacities of paid human labor, we will be like horses, inferior and expensive.
4
UBI will never get passed in Congress without elimination of food stamps, housing subsidies, ACA premium subsidies, the earned income tax credit and any other assistance to poor citizens. Rational, "we gave them cash let them decide how to spend it". Also it won't be indexed to inflation - minimum wage isn't so don't expect UBI to be.
UBI will never become a reality until we have "a revolution!"
2
Automation doesn't just reduce the amount of labor required for a product, it also reduces the product's price. This leaves more money to the consumer, which stimulates demand for new products. Computers, restaurants, gyms, dental implants, are all modern examples. As long as enough money keeps going to the largest number of consumers then we shouldn't have to worry. Unfortunately we have been keeping wages low so that a few of us can become very wealthy, this is the real problem.
4
So where I work we just let go 75% of our website testing team because automated tools testing code for usability and security have gotten so much better.
With people living longer and having to work longer Krugman here misses the people problem of what happens when a person loses their job in their 50s due to automation. Retraining for another job is one possibility sure, so is depression and addiction.
12
@ahimsa Not to mention if you are in your 40s with a spouse and kids and you lose your job, you can't just retrain or start over, you need a job to support your family asap. And no one wants to spend a 5-10 years going to college or retraining for a new career only to find the job is gone when they are done. Constantly retraining is a poor answer.
If tech jobs are getting automated, that's scary.
4
Krugman misses the point when he says increased worker productivity is a measure of automation taking away jobs. What misses is that it's the pursuit of Reduction in COST of production that matters. That's why manufacturing jobs moved to China in the first place, and are being replaced by automation. It's also taking away service jobs, as exemplified by Yang with automation of truck driving. Why is GM closing plants? What data is Krugman sharing of what's happening right now? What UBI is supposed to do is not a way to scare people about a dystopian jobless future, but give them breathing space to gain new skills for a changing jobs dynamics, and maybe even better family and social interactions than the ones we have at present.
5
This is Krugman’s second op-ed (that I’m aware of) that has the intention of undermining candidate Yang by stating the threat of automation is overblown by his campaign.
In both pieces he states not many jobs will go away, and it will take a long time, however, he misses the gravity of the situation.
Countless studies (The MIT technology review has a great summary of these studies) points to millions of jobs lost, jobs that will never come back. Now with AI or even just computer algorithms professions like counseling and anesthesiology may be on the chopping block.
What Krugman does get right is that there’s a lot of work to be done in this country — and globally. The answer to this though, is not to ignore automation, but instead embrace it, and leverage it’s efficiency and profitability to lift up everyone’s quality of life. Yang’s answer of universal basic income is a great jump start to this process.
24
@Nathan Corliss : Until recently globalism was a saint and the way forward in Krugman-land.
@Nathan Except $1000 a month isn’t much of an income.
It'd be hard to call Nobel prize winners wrong but each robotic arm in automobile industry replaced 6 workers which can be linked to the area code of where the robotic arms were deployed. Yes, it happened in Detroit. If it created some jobs for making the robotic arms, it happened elsewhere, not in Midwest. So, as soon as the AI wave combined with automation becomes profitable, many jobs are again going to be extinct. Without sounding apocalyptic, future of work does not look good. It's not good for my toddler when he's 32 in 2050. I'm sure that new models will emerge, but it is part of the democratic process to discuss and debate future of work.
12
@Sunny :"Nobel prize winners "
Krugman is not a Nobel prize winner. There is no Nobel Prize for economics.
It's called: The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel.
Food for thought and room for debate.
What's with the big switch? First time this column ever said anything remotely positve about a Bernie Sanders idea.
Sanders never got the basic respect of evidence to refute his ideas. Just say 'unworkable'.
Sanders -- whether or not he'd be the best to win -- was a fearless truthteller. The establishment tried to reject him, while still looking better than the GOP.
Many liberals put down Sanders and any true progressive that could right our wrongs. It's even simlar to how the GOP insults all progressives. Just less vicious.
2
I fail to see how the rate of productivity gains disproves the existence of adverse effects from automation. It seems like a lazy argument, and I wish he would try harder. In the past, it was easier to find more productive jobs as innovation displaced workers. These days, it seems like more people drop out of the work force or end up as minimally productive gig employees. If we really could retrain retail workers to become coders, then maybe we would see productivity skyrocket. But we can’t, and that’s pretty much the point.
Yang’s reaction, in my opinion, proves that Americans do feel impacted by automation. Krugman’s interpretation of this as escapism strikes me as overly academic and out of touch. Elite liberalism at its finest.
16
Worker productivity rose half as much from 2007 to 2018 than in the previous eleven years. And what debuted in 2007? The iphone. I suspect that distraction from cell phones leads to significant productivity losses. (Look around your workplace.) Employers will be happy to replace humans addicted to their small bright screens with robots who’ll get the job done.
4
A couple of things should be obvious to all of us. Every society which has a core of scientists and a capitalistic economy will see industrial like revolutions. History has also shown that as societies we have always benefited from this combination. Lastly, these revolutions can not be stopped. Life has only gotten better with a combination of science and capitalism.
2
Thanks to another reader, I bothered to read Yang's UBI platform. This is a pretty exciting and winnable idea. Why isn't everyone talking about it? There seems to be real evidence that it works, it's income-blind (so no complaints of Robinhoodism) and doesn't contradict capitalist principles. Is it an issue of too simple to understand/sell?
21
And yet it's safe to say that whatever can be roboticised will be roboticised, depending on relative costs and benefits of each option - robot or human. Up to now, robots have been used for rote or repetitive activity. What will be new is robots moving into service industries with the advancement of artificial intelligence and web-based services. History shows that everywhere a human can be replaced by a machine, it has been.
Also, as corporations pursue their ultimate purpose, to produce goods and services at the lowest possible cost, the cost of labour will be forced even further down. The political environment is allowing that to happen through relaxed workplace protections, labour laws, tax policy, and, yes, trade agreements. Once all the money that can be is squeezed from the cost of raw materials and capital, they will hit labour costs hard. Today there is no political will to stop it.
People who are in many of the jobs to be lost will not be gaining employment in high skill tech jobs like designing and repairing those robots. The expectation that everyone will be so qualified defies the reality of the direction in which public education policy and investment is going in the US.
So exactly where will investment in the US go, to people or to further automation? And what will be the end result of those investments? I believe it will be more about the latter than the former. I'm glad I won't be in that workforce.
18
"So harping on the dangers of automation, while it may sound tough-minded, is in practice a sort of escapist fantasy for centrists who don’t want to confront truly hard questions. And progressives like Warren and Sanders who reject technological determinism and face up to the political roots of our problems are, on this issue at least, the actual hardheaded realists in the room."
Yes.
2
@Julie :Here's Warren's take on the GM strike:
“Their only loyalty is to their own bottom line. And if they can save a nickel by moving a job to Mexico or to Asia or to anywhere else on this planet, they will do it.” (NYT, Sept. 22, 2019)
Krugman must be covering his ears with his hands saying "I can't hear you!!!"
People need jobs. We have roads, airports, schools, public spaces, etc., all over the country that are in desperate need of repairs, upgrading, or even tearing down and rebuilding. Some of that can be done using automation. But it takes people to plan, people to coordinate the tasks, and people to do the tasks computers can't do. The longer this country refuses to allocate money for our infrastructure, to get us off of fossil fuel, to preserve our national parks, our water, etc., the sooner the country will become uninhabitable for most of us.
Jobs have changed. But there are new jobs waiting that require a human touch, human judgement, human knowledge. Businesses and our government are wrong when they refuse to invest in people and the more they refuse the more this country will fall behind the rest of the world. We will become a pariah and one of those countries Trump denigrated. But it will be our fault because of who we vote into power.
Elections matter. Be careful who you vote for and think of what will benefit everyone, not just you. We are all in this together.
10/17/2019 9:44 first submit
11
@hen3ry I actually disagree. People need the means to support themselves and their families, and to make their lives better. Jobs are something that we are willing to do to achieve that. It's not the job that we need, it's the remuneration. (I will acknowledge, however, a need to feel useful.)
To the extent that automation reduces the need for human labor, if the benefits are distributed equitably, that should improve our lives and our society--giving us more time to raise our children, care for our families and our neighbors, and contribute to our nation.
It turns out, however, we have decided to build up our military and protect massive wealth rather than make investments in things that improve lives for everyone.
We can always say there is infinite demand for good and services: "Built it and they will come attitude".
It is not a fix rule of the universe like entropy increasing though.
So putting our faith that massive automation will not create massive job market disruption is lacking vision.
There are resources that are non renewable or renewable at a slower rate than we consume them, infinite value added growth is kind of missing this point.
In economies with aging populations for example, consumption of goods is likely to drop.
There also all those artificial job creating activities like pushing 600 Billions into military spending with little return on investment from a added safety prospective. It does create millions of jobs though, will those shift ?
So while I hear you about not running down the robot rabbit hole, there still a long term discussion to be had on the topic of future of jobs.
Infrastructure work do not create as much jobs has they use to. That pretty highly mechanized, not like building condos [they will be 3 printed eventually too] .
We should have massive investment in renewable /sustainable energy, urbanism, housing … But those jobs are not for everyone.
Lets say you wanted to install solar panels on 50 millions houses at 6 person*day per installation (2days, 3 guys), that 300 millions days of work or a year of work for one million people. Year two, you would run reach market saturation. Anyhow, it will not replace 10+ millions jobs in transport.
4
It’s not about automation as much as it is about one thing; money.
There are plenty of things that need doing - climate change alone is the big existential threat. Rebuilding our infrastructure to adapt to it and get us off fossil fuels would create jobs all over the country. That would call for massive amounts of public and private spending, putting money right back in the hands of working people.
It’s not happening because Capital is all tied up playing games with money to make more money. There’s more profit in that than there is in saving the planet.
This is one reason we need a wealth tax- to get that concentrated wealth spread out to where it can do useful work for real needs.
12
The first law of cybernetics will thwart the arrival of self-driving vehicles in our lifetimes. Pilots and truck drivers will still be needed for some considerable time to come.
@CC No doubt, but larger passenger planes have already been automated to the point that various parties have expressed concerns about the declining skills of pilots as a result of doing far less actual flying. The planes are on autopilot much if not most of the time, and pilots are increasingly unprepared to act when emergencies arise, since they've been conditioned to rely on autopilot.
And many other jobs WILL be automated.
2
@Stratman Just because a plane has an autopilot doesn't mean the pilot has to use it, especially on VFR days. Any pilot that allows his/her skills to slip is not using the automation correctly.
1
Aside from the hyperbole 're' automation concerns, you may have noticed no one, I mean no one, spoke about the need to control a 'universal storm' already biting our 'behinds', Climate Change, as we are the main contributors to the increased frequency and severity of natural disasters. And without doing something to reverse the harm, all else may be just an afterthought. You mentioned inequality, which is more complex than just the economy, as it has racial and gender components as well...plus the institutionalized segregation still rampant, be it in jobs and housing, or in health care and education. And further, the lack of healthy private corporate competition, as oligopolies are allowed to create havoc to smaller -yet as valuable- companies, suggesting we have replaced ethics for selfishness and greed, rendering capitalism an 'enemy' of this democracy. Can't we see that our society abhors a vacuum, and Trump's ability to cash on it, to fill it with his superb demagoguery, a disaster for America... forcing us to re-invent ourselves to integrate the least among us in this 'adventure', the restoration of trust in our institutions?
2
Many comments from neo-Luddites. Stasis in not possible; change is inevitable. No one (or, hardly no one) morns the passing of harness makers and horseshoe fabricators from the economic scene. People must be flexible in their job skills and, above all, be willing to relocate.
Dr. Krugman's analysis is dead on the money.
3
There are a multitude of dynamics, now taking place. Our economy is quite complex and has been buoyed up by the cheap money policies Greenspan put in play, a decade ago.
Anecdotal evidence. The teller at my wife's bank was replaced by a video monitor. Not all, but many transactions could be made that way. I don't think self driving cars and trucks are far in the future. Eventually they will be safer than humans.
Trade wars and income inequality are very real. Bankrupt pension plans are soon to be. Negative interest rates may be the only answer, but that hurts many.
If Trump or Warren are elected in 2020, please advise on relocation possibilities.
3
@Texan
Canada has a more affordable healthcare.
But a lot of the economy is still connected to oil and trading with the US, so if the US get a cold Canada gets pneumonia.
Still comparing Warren with Trump is kind of silly. Warren is a lot smarter than Trump and will not start a trade war.
I will pay more taxes, wealth tax or inheritance taxes... But that Ok with me if this help making the US a less unequal country with less child hunger.
Sustainable wealth (shared equity) should be our goal because wealth is a zero sum game. When the rich get richer the rest ALWAYS get poorer. We don’t need economists guiding us. We experience deregulated capitalism’s inequality every day.
4
Automation is why there have been job losses in manufacturing. It can't have been trade, because the total dollar value of our manufacturing output has been continually increasing (except for a dip in the Great Recession). If manufacturing was moving overseas, our output would be falling along with jobs. But in reality, we're producing more stuff than ever before, just with fewer workers.
Automation hasn't affected service jobs as much--yet. Lower productivity overall could well be because workers are being shifted into these lower-productivity but harder-to-automate low-skill service jobs. But it's certainly a possibility in the future, and no one knows when it will happen. Self-driving cars alone would take out millions of driver jobs.
Of course, this automation will be a good thing for society overall because we will be producing more--and jobs exist to produce goods and services that we want to consume; they are not an end in themselves. But there will need to be some way to share the gains. A UBI could be a good way to do this. It could start at a low level today, and gradually increase as our GDP increases due to automation.
2
The self checkout at the grocery store constantly malfunctions. The computer screams when an item isn’t placed in the bagging area fast enough. It scans things at the wrong prices. It demands that people wait for assistance from the long suffering human employee who is frequently gone from the area so customers end up waiting longer than it would have taken to go through a line with a human checker. They went to self checkouts because they couldn’t fill enough positions with humans, probably because they refuse to pay a living wage.
715
@Bookworm8571 Exactly. I am always bemused by the number of staff hanging around the self-checkouts to actually complete the purchase. However, the machines will improve, but they will also make mistakes without having understanding of what a mistake even is (at least until the Singularity) which potentially will lead to catastrophic glitches. At it is you, as a human, can never win against the voice-operated help call lines!
39
@Bookworm8571
Self checkouts will never work well until the entire store is redesigned around a new automated checkout system. It is uncertain whether all the pieces of technology needed for that system yet exist.
11
@Bookworm8571
I worked in grocery about 1970 or so when I was in high school. We rang every item up by hand. Scanners came and I watched cashiers scan and it came up to speed but nobody knew what anything was anymore; especially produce. Now when I scan at Fry's I talk back to the bossy machine but make less mistakes than the paid cashiers. In fact it's so convoluted that they have to pay somebody anyway to keep the machine happy.
26
The difference is that AI will soon allow for a far greater level of automation than before. It's very easy to foresee a future not to far away where AI controlled devices make large number of workers obsolete, and not just in blue collar jobs, but in white collar jobs as well.
That doesn't mean we should run around screaming the sky is falling, but we should treat the issue seriously. It's pretty conceivable that we could have a small number of very rich people that don't do any work, and a large number of very poor people also not doing any work, all the while the machines hum along.
Think about a small subset such as autonomous driving. What happens when millions of drivers (semi's, taxi, delivery, uber etc )are no longer needed? We should think about these problems now.
350
@Mathew "It's pretty conceivable that we could have a small number of very rich people that don't do any work". You mean even more than now?
39
@Mathew So why should the machines be "humming along" when there's no market for what they produce? How can the idle, but very poor, people buy much of anything without income-producing jobs? Get rid of all the workers, and every industry loses most of the reason to make products. Because they won't sell.
61
@Mathew -- "It's very easy to foresee a future not to far away where AI controlled devices make large number of workers obsolete, and not just in blue collar jobs, but in white collar jobs as well."
It seems to me that it would be bad 'business strategy' to get rid of all your potential customers. Who buys the products of these automated factories, when the jobs for humans are all gone away?
26
Yang is a single "solution" (not issue) candidate who has few ideas outside of his UBI scheme. I wish the debate moderators would have asked him about foreign policy, climate change, gun violence, or immigration. In my view, he is running just for attention and detracts from crucial debate time.
2
@LonghornSF go to his website, he literally has over 80 policy proposals. I wish they would have asked him this questions as well. You would have heard that he wants to exchange guns that are only usable by the owner thereby decreasing mass shootings where guns used were taken from the owner like a parent. Or that he acknowledged that climate change is worse that what other candidates discuss and these simple solutions won't cut it. But the thing I like.most about him is his dedication to a single policy, one that would help more red States and thereby be easier to pass into law than anything any other candidate offers. In addition, if I vote for him I know what he wants to do starting on day one. Maybe except for Bernie with mfa and gabbard with ending intervention, I don't know what anyone else will use their little political capital once they get elected.
2
Automation in the past wasn't outpacing people's ability to adapt to the transformation. Today, technology is outpacing our ability to utilize it in a way that adds meaningful value to our work. One can be extremely intelligent but not be cognitively wired to master let alone cope with the advances that result in upward mobility, which is the real driver behind the increasing income gap.
3
Dr. Krugman, you should have a beer with Thomas Friedman, that most vocal, and disingenuous proponent of "lifelong learning". "Lifelong learning" is the plutocrat's answer to chronically low wages and mass dislocation, putting the entire onus, for success or more likely failure, on the worker. Friedman urges us to devote our lives to the Sysiphean task of remaining a useful widget, until finally, out of exhaustion, we collapse and die. There are several reasons why this is pure foolishness. Most people past a certain age are neither equipped or interested in continuous learning, and that applies even to the things they are supposedly interested in, much less the numbing, pointless tasks they are paid to do. Whenever the same experts discuss the "jobs of tomorrow", the question arises as to where we should allocate our resources now to prepare, but they can neither tell you what those jobs will be or what will be necessary - trust them though, the jobs will be there!. Lastly, on-the-job-training has disappeared from the landscape. Employers expect you to show up fully formed- like Athena out of Zeus's brain. Of course, this is also a thinly disguised ruse to hire cheap foreign labor, and once an employer finds someone willing to do the work for one third the wages, suddenly that list of impossible job requirements magically evaporates.
6
I feel like this time may be different because average human intelligence is not increasing, while computer technology gets more advanced every year.
We know that human IQ is quantifiable, and difficult to increase, so the question is if we are getting to the point where half or more of the population simply won't be able to handle the rigorous education required to work in the jobs that will supposedly be created by AI and automation. If only 20% of the population is intelligent enough to handle the kinds of jobs that will be created, then Andrew Yang has a point.
10
I am an early Yang supporter, but it's not because I care about automation vs trade in particular.
The reason I enthusiastically support Yang is because he is bringing UBI to the political discussion. And the reason I care about UBI is because I believe that poverty is profoundly expensive to individuals and society. MUCH MUCH more costly than we collectively acknowledge. UBI is the best solution to poverty I am aware of.
Trump won the last election in part by using immigrants as a scape goat. Even if you don't think automation is the root of our problems, we should recognize that automation is an excellent way to divert the subject of the frustration felt by Trump's base. Just look at how many former Trump supporters are switching to Yang! Automation is as much a political strategy as a policy focus. I fear the author misses this point.
8
It's all about the "Social Environment" in the C-Suites; We now have two generations of number people (aka "MBA's") who are convinced that Labor is a cost, and he simple way to boost the bottom line is to get rid of those pesky Union (Fill in the blank). The "Lump of labor" theory (Except in the executive suite...). An Indian code contractor is as good as in house specialists (Boeing); etc.
They engage in anti-Union tactics that would have resulted in Social Ostracism and Congressional Investigations in the 1950's; But they have captured the law writing process, and made it all legal.
Scheduling software is as much about control of the work force as costs; It results in an anonymized work force who lack the social connections and trust to even talk about the U-Word.
Rather than raise the pay and think about better scheduling practices fro Truck Drivers, they wanted to import (presumably lower wage) Mexican Truck Drivers.
What are we going to do with people who can't keep up with the computer driven schedules at the Amazon warehouse, etc? And again, unionizing those work places is an uphill struggle (see paragraph two, above).
It's not the robots, it's the MBA's.
3
@OzarkOrc Uh, labor IS a cost.
What would really be interesting is if Krugman would explain why there has not been a rise in productivity among American workers like that to which we had become accustomed. Might it be that the big tech innovations are kind of worthless? How does Facebook help productivity? Harder question: How does Google? What productivity do we get from Ebay, Uber, airbnb, twitter, instagram, or any other of the headline web giants. What about Amazon? Seems like their innovation was in fast delivery and low-paid workers, as well as strike-breaking. It makes me think the biggest improvement that the web has given us is email, and the online news (but real journalism, not facebook or aggregator news).
294
@Daniel Mozes
Well, the short answer is that Facebook, Google, Ebay, Uber, and their ilk are not tech innovations, as pointed out in Jackson's comment just above yours. For tech innovations, look to things like GPS, perhaps someday quantum computers, microprocessors, cell phones (even though they are nothing more than two-way radios), firmware (like in your car), the transistor and integrated circuits and other more recent shrinkage of components, the internet and so on. Those are hardware with software components, but for pure software, consider Gary Kildall's OS CP/M (from which DOS was stolen, a story well documented and often told) as well as the bios, which he developed, FORTRAN, word processing on PCs (which put a bunch of people out of work, the so-called Wangers who in large numbers packed rooms creating documents on special purpose machines made by Wang Laboratories which were a new technology that changed the office place), spread spectrum communications, the digital phase-locked loop to name a very few.
38
@Daniel Mozes I'm a researcher. Trust me, Google has revolutionized the ability for us to quickly find information, even compared to other online sources like Pubmed. What takes minutes now to Google may take me several hours per day researching in the library in hard copy journals.
65
@Daniel Mozes As a custom woodworker, Google, Amazon, and the like have tremendously increased my productivity, as they have for my competitors as well. Thus there has been no gain for us, but our customers have benefitted from downward pressure on pricing.
40
Automation and AI is good in relieving people from doing menial and even complex task. It won't be the end of the world if people no longer have to drive trucks for hours on end, for instance. We'll just have to find other harmonious and empathetic ways to share in the prosperity.
3
The difference now is that the displaced jobs aren’t to be replaced by other lines of work requiring growing numbers of employees that offset the changes. People aren’t leaving agriculture for factory jobs as they once did. The ten self-checkouts in Walmart’s Levittown retail lab haven’t led to ten tech jobs to install, maintain and repair them. The Levittown Walmart Neighborhood Market replaced a unionized supermarket with attended checkouts. The cashier positions it is displacing are being absorbed in the economy by part time and gig work at lower pay and with fewer if any benefits.
470
@From Where I Sit This is just anecdotal - you need some data to make a compelling point.
4
@From Where I Sit
In your own example of self-checkouts, work has actually not been eliminated (someone is still scanning and bagging), it has just been shifted from store employees to customers. Are we willing to accept the same outcome of automation in health service industry, though?
21
@From Where I Sit There are always changes coming that open up new jobs. Whether people want to admit it or not, fossil fuels are on their way out as an energy source. This country will be replacing its entire energy system with one based upon a panoply of renewables. This will necessitate a completely different power grid allowing a much greater variety of inputs. It won't be robots building and installing this enormous new infrastructure.
17
One need only watch truck drivers in the next ten to twenty years to confirm the seriousness of the automation problem. However, automation fears and inequality are tied together. People are not afraid of robots, they are afraid of robots replacing them and not seeing any of the profit from that automation. We cannot stop automation, so we must guarantee the profits from its implementation are shared amongst those it is replacing.
467
@Jackson
Yes, this is exactly the problem. In a free market, it is reasonable to expect every participant in the economy to seek out anything that increases profits. Clearly a robot that can do the work of 50 human beings, never needs to take a break or strike for higher wages and is a great deal for stockholders, less so for the 50 replaced workers.
Also, we can assume that jobs robots can't do, many people are not smart enough to learn. Some people will swim like fish in this brave new world, but many will sink into depression and self-loathing as they fail to keep up with rapid change. This is already happening in Trump country.
The intelligence required to keep ahead of the high-tech curve is not doled out equally by Mother Nature. It used to be that the less luminous minds among us could still get by well enough to permit normal lives, including marriage, children, home ownership, a bit of leisure time and a modest retirement. Being willing to work hard was as important as being smart. That is now a thing of the past.
So to permit the people who drive, flip burgers, bag groceries, sell stuff at shopping malls, collect bridge tolls and do a million other useful but not mentally taxing jobs that - sooner or later - will either be done by robots or pay people less than a marginally living wage, the wealth will have to be redistributed, and the details of this will have to be determined democratically.
Inadequately regulated capitalism is not getting us there.
54
@David in Le Marche Exactly what Yang is addressing with his guaranteed income plan.
15
@David in Le Marche
You nailed it. And so clearly stated.
Warren blames it on oligarchic capitalism but that's just her creating an enemy to fight.
The truth is when K-12 public education includes technology skills and the achievement gap between rich and poor, black and white, girls and boys, rural and urban is gone, America will be okay again.
5
Robotic automation will continue to replace workers in manufacturing. Outsourcing workers to far-away countries with lower wages will continue. Most lower skilled jobs have already been shipped off shore (think call center employees). Higher level skill sets are now being shipped off shore (think architecture and legal services). Many lower skilled service jobs in the US have turned into 'gig' jobs with no benefits. The higher level service jobs that still remain require higher productivity from workers to maintain their status and benefits. 40 years of vulture capitalism has driven productivity and cost reductions to an extreme - and extracted gigantic profits for corporations and the 1%. The 'free' market will not fix this.
5
I often agree with Paul Krugman’s economic take. But this time, I take exception. As a physicist and someone who has also worked in the policy arena for years, I have a pretty good feeling for revolutionary technological change. Yes, as a nation, we have weathered, benefited and generally prospered during every technological game-changing era. But the changes came incrementally. so our population was able to adapt over the course of one or two generations, by and large without government meddling.The pace of change today is much faster, and without programs to help people learn new skills and cope with geographic upheaval, we will sow the seeds for ever-increasing income and wealth disparity.
As a professor, I take issue with the philosophy dejour that argues higher education should emphasize programs that train students for the job market, so they will be optimally employable when they receive their diploma. For an era where the jobs of tomorrow will not be the jobs of today, we would be far wiser to emphasize learning skills rather than job skills. Lifelong learning will be de rigueur in 21st-century America. We’d better get used to it and be prepared.
874
@Michael Lubell
In addition to rapid technological change and global competition, the political landscape is also being reshaped. Intelligent conservative political thinking is being idled and replaced by political influence from overseas. Masquerading as the people’s champion is cheap imitation self-interest, made in Russia.
53
@Michael Lubell : Might I add, today's technological changes don't add jobs like the past. Google is worth billions, but what percentage of the population does it employ? Compare that to the auto industry and its supporting vendors.
I do agree with Dr. K in that as a nation we need a massive influx of infrastructure spending. We have constantly cut costs, which meant needed maintenance and replacement needs went unmet.
83
@Michael Lubell but for lifelong learning to be possible this country has to start to respect education, intelligence, curiosity, and all that that means. Very few Americans understand science or how it works. Even fewer care. If you try to explain to the average American how and why theories change they claim it's too difficult. What we need to do is overhaul our educational system and teach critical thinking instead of memorization and agreement.
Government can help but the people we put in power won't. What we need and what they do are in conflict with one another. Until we decide that we want a functioning government and one that serves us we are not going to see much change in our country.
124
I keep waiting for someone to talk about infrastructure as
part of a “maintenance economy”. What Mr. Sanders calls “jobs to be had” and “work to be done” is essentially maintaining and upgrading what we already have. It is part of the larger economy in that it sustains our labor force, but does not rely upon growth to the exclusion of sustenance.
My mother always said that 90% of life is maintenance, whether brushing your teeth, paying your mortgage, training your kids to replace you, recovering from natural disaster, voting, or paying taxes to sustain our infrastructure of roads etc., schools, courts, environment, and health.
Mr. Krugman, how much of our economy is embodied by maintaining what we already have? 90%? Mightn’t we redefine taxation? ...not as a penalty upon private wealth, but as a maintenance/usage fee for our societal infrastructure: our common wealth? The more private wealth you earn within the web of our societal infrastructure, the more you implicitly owe back to the society that provides it.
5
I agree that automation in general doesn’t eliminate many jobs, it jus moves them from one place to other, with new jobs usually requiring higher education. In agriculture fewer workers can manage large farms because of machinery but this machinery required many more qualified workers to design and manufacture. More materials were needed, more miners were needed, etc. Price of food dropped but price of other materials went up.
One way to to manage this process may be to increase the value of manual labor and, for example in agriculture, change they way we manage arable land. Higher wages in agriculture would increase the price of food and would make smaller and environmentally sustainable farms economically viable. Many young people would gladly return to work the land if there was economic stability in this sector. Higher food prices would also have other benefits like reducing the food waste that is estimated at 40% now. Capitalism is enormously wasteful system and we are slowly drowning in its byproducts, waste and pollution. It is simply unsustainable.
Of course these changes would require significant intervention from the government and require strong regulation of capitalism. Though I think this is simply inevitable, we have to deal with waste eventually, I fear the ignorant, greedy and selfish public may be not be there yet.
1
While I can see both sides of the "automation" argument, one of the things we keep missing is that these changes are significantly due to what consumers demand.
We, the consumers, push on the other side to improve speed, quality, and quantity in just about every area.
When you couple that with the fact that companies have the leverage to just move to some other place to contain their spending, well, that leads to some imbalance.
If I had to guess I would say that at least 65% of us responding to this article, whether we agree or disagree with Paul Krugman, are "Prime" members at Amazon, a company that is taking the road of "automation." But we still love Amazon.
At the end of the day, it all comes down to how balance the game is.
Companies have the leverage to move abroad to some place where labor is cheaper. We don't have that much leverage.
I think Paul Krugman capped it off in this reference to "economic weakness."
"[T]he persistent weakness isn't about automation; it's about inadequate private spending."
1
I agree with Krugman. The last time I talked with a robot on the phone I could not get to a live person fast enough. The companies that don't let me get to a live person won't get my vote (purchase).
2
My dream in the early '60s was that as technology developed workers would have a 3 day work week and a livable income. Time to pursue, art, writing and community service. Even time to be a stay at home mother.
6
Hey, why not admit that economists know nothing about how to control, start, and manage economic growth. And this reader is not buying your take on trade policy with China and your dismissal of the possibilities of manufacturing here. Economists like you let "free trade" destroy the Rustbelt over a 40 year period, pretending there was nothing that might be done and that the winners "could theoretically" compensate the losers so that all was good. It wasn't good. You bear a lot of the blame for what happened, which includes Trump winning the White House.
2
Paul clearly does not spend much time in engineering and tech circles. The fastest growing job in the planet, Computer Science, is designed to replace humans. The real fantasy is believing that humans can out compete computers.
7
I agree. The blather that so-called artificial intelligence will lead to mass unemployment in white collar jobs is just that, blather with no substance. Data processing, which is and always will be what CPUs do, remains data processing. There is no "intelligence" involved in the matter, at least intelligence as defined by human or even animal cogitation. Billions of years of development of such intelligence through live-or-die experimentation are not overcome within a relative split second.
That more data are processed faster leads to possibly better tools for mankind to use. Possibly because so far, the mass adoption of mobile computing devices has probably had a net zero or possibly net negative impact on human productivity.
The analogy of data processing to one of the most primitive of mankind's tools, the hammer, in one form or another, holds. Faster data processing is in effect a more sophisticated hammer, but still a hammer.
2
Vonnegut’s Player Piano is a great book, but it is not only about technology, it is also about the value that we place on intelligence as measured by tests and diplomas. It is about a society that values only that which can be achieved with math. It is humanless and ultimately soul destroying. A dystopia that is what a focus on STEM leads to. It is what the pin on Andrew Yang’s lapel foreshadows. A world of engineers and managers. A world of gray autonomous living death.
6
Sorry Paul, but I think you are confined here by a temptation to normalize the disruptive capabilities of technology. The iterative development of technology has changed the entirety of our world and culture in just the past quarter century. Dismissing reasonable thoughts about the potential impact of capabilities in their nascent stages out of hand with a reference to the mid 20th century is, well, kind of absurd.
It takes an understanding of what these capabilities will be. Specifically the unprecedented potential to to replace expensive and, relative to machines, unreliable human output is quite real. Impacts will range from trucking (which alone represents a massively disruptive shift) to the previously untouched information services industry.
Yes, there are real and ‘boots on the ground’ opportunities to put people to work. Sustainable energy and infrastructure, to name two, that could put a generation to work if lawmakers could get over control battles and get to work on a roadmap for our future. But to respond to the very real emergence of technological changes that will dwarf the digital revolution in this manner is dangerous. We need an ethical perspective focused on the primacy of humanistic values before fiscal bottom lines if we are to avoid an Player Piano future.
7
Mr Krugman seems not to recognize the massive job cut in his very own industry, if not based on automation, than certainly on its sister digital transformation. The reason for Andrew Yangs success is that people working in the digital industry know what’s coming, and people in the Midwest Starting being effected by that change.
7
Paul, I would love to see you have a public debate on automation with Yang so that we as workers and voters could weigh the evidence and arguments for your different points of view on this important subject.
14
This is a real problem with the Uber nerd tech genius types. I KNOW what I’m talking about, with 40 years of marriage and counting, to an Engineer. Very high IQ, little emotional intelligence OR common sense. The best way to deal with it :
Just smile and do what you want. And Wine.
20
Car factories used to have thousands of employees in the glorious 1950's. Now I see robots with big arms doing just about everything and maybe 200 technicians making sure all the robots are running smoothly. I am with Yang on this and I think professors Krugman and Warren need to look beyond "the data".
9
Nah Paul, You've not got it either. It was the combination of computerization of the workplace AND shipping of jobs out of the country via NAFTA and trade detent with China, (Clinton). I don't agree with Mr Yang's solution and I don't agree with Ms Warren's explanation. Ive lived through it. I traveled to Japan in the early 80's with the chief engineer shopping for then best in world, Japanese made, CNC production equipment to modernize our factories. Productivity is output per worker. If you send workers home via computerization of the factory or farm, productivity as currently measured goes up, cuz output improves with less workers. Not getting this is why we had the Tea Party and have Trump. We excelled economically in the 50s as you describe cause we were the industrial nation left standing after WWII. We put people to work feeding and supplying stuff to the world. The world recovered.
6
One of the few times I disagree with Mr. Krugman.
I think Yang is more perceptive of, and better equipped than other Democratic presidential candidates to deal with, the accelerating replacement of human skills by machines.Not going to argue for his point of view here, but you can look at recent trends in careers such as paralegals and radiologists for examples.
My 2020 dream ticket would be Senator Warren as President for her deep experience and understanding of our national system of disaster capitalism and its remedies, with Andrew Yang as Vice President for his understanding of the rapidity with which automation is changing our world's economics and contributing to wealth concentration.
5
It is not wrong to predict that AI and automation will have dramatic effects on the economy. However, that is not what PK is talking about. The 2020 presidential election will determine policy for the next four to eight years. Highways will not be full of self-driving trucks on that time scale.
We should certainly prepare for the longer term, but that is not the issue at hand.
3
Krugman: because we’ve faced technological watersheds threatening jobs in the past, this one will be exactly like all the others.
But to paraphrase Yuval Noah Harari, being exploited is different than being irrelevant. Expect to see most of the uneducated masses of the US electorate moved into the latter category.
4
THANK YOU Dr. Krugman.
I retired after 40 years in IT, the last 23 spent working overseas for large companies such as Johnson and Johnson and Pepsi-Cola.
Automation does not kill jobs, it creates jobs because it creates productivity. One of the most frustrating things I dealt with in business was the lies told by the (non-technical) management, that a new system would save money by cutting jobs. Once the system was installed, inevitably there was a need for more, different kinds of jobs to run it!
I saw two problems over and over: senior management dishonesty driven by greed and ignorance, and moving jobs overseas or importing cheap (and always worse) workers to save money in the short term.
Tariffs are not the answer. The answer is to prevent imports from countries where they can pay slave-like wages and pollute at will. The TPP was a giant step in the right direction, and we should return to that approach that uses strong alliances of non-corrupt democratic countries to enforce fair trade.
I am a Democrat, and I dearly hope the party will exclude the likes of Yang from all future debates. He is adding major amounts of disinformation to the campaign.
27
@Terro O’Brien
Why does no one ever point out that most managers could be replaced and we don't even need AI. That would also solve the problem of too many stray dogs in shelters.
2
Krugman is presenting only one side of the coin. Overall labor productivity has not magically increased, but it has increased, for the overall workforce. This does not negate what automation has done to manufacturing, which is what Andrew Yang primarily discusses, particularly in the manufacturing swing states.
To see this for yourselves and understand why Krugman's data is misleading, compare the number of jobs in manufacturing, which has declined from 18 million in 2000 to about 13 million today (https://fred.stlouisfed.org/series/MANEMP) while real output has increased (https://fred.stlouisfed.org/series/OUTMS). Please explain that, Mr Krugman, and also explain the opioid epidemic, the stagnant income growth, and the declining life expectancy in this country.
Also consider how automation has not yet affected the service industry widely--cars that drive themselves, computer systems that can read and write like humans (even talk and perform tasks), diagnose like physicians, spot patterns, and learn from massive amounts of data, etc. This and more are just around the corner and will take away most menial and repetitive physical (and cognitive) tasks.
Automation will prove extremely beneficial for all if the spoils can be shared to more people, but to say, don't worry, it's nothing we haven't seen before, is incredibly naive and shortsighted. Also, please remember how disruptive the actual industrial revolution itself was.
18
@T well said. With the advances in robotics, AI and quantum computing, we (none of us) know what’s in store, for good or/and I’ll. Paul, I love your columns and you’re far smarter and knowledgeable on economics than I, but I can’t help thinking you’re you might be underestimating the next ‘revolution’.
3
@T
I am sure there is no replacement for for human intelligence, in diagnosis, healing, geo-vitalization, and things unheard of such as huge corrections in our relativity to life, consciousness and each other.
The problem and danger that I see is loss of human capacity, which we gained through our evolution with all life forms over eons. Automation will be a crisis in human identity but will lead to identifying awesome levels of human capacity which have been considered impossible. These are happening already and not because tech is the solution. To believe otherwise is just plain boring, which is why we - or I, don't consider tech superior in any substantive way. All the numb aspects of human imprinting are what is creating tech, while aliveness is irreplaceable and exponential.
But there is a necessary divide happening; My esoteric take.
1
I’m no economist but I’m pretty sure it’s wrong to so completely discount the probability that technology will more and more render human effort unnecessary to produce all of the things we need and all the services we require.
The fact that labor adapted to other technological changes in the past is unconvincing to me. We might have thrived in the post war years, but eventually we settled into the long term trend, I believe, which is the decline in value of human effort, at least the kind of effort that most of us have the capacity to offer the world.
We need to, eventually, sever the notion that one has to earn one’s place at the table by hard work. And we have to start valuing being human. Krugman might seem right in the near and medium term, but he is obviously wrong about the the long haul.
7
Thank god for Dr. K setting the record straight. Two heads are better than one.
I was already panicking about future shock, the rise of the machines, where I, robot, rules.
2
I am so happy that Dr. Krugman identified Erin Burnett's question as a bad question. I would add that most of the questions were bad, from all three moderators. Because they wanted rebuttals, for entertainment reasons, they often trotted out falsehoods. The candidates then had to explain -- as if you can in 45 seconds -- what the original statement was and took time away from each and everyone of them being able to talk about policy. I understand that network TV and politics are combat sports, but this was degrading.
9
You are absolutely right. Innovation makes more jobs. In China, India, Mexico and other places where the cost of labor is cheaper than the cost of automation.
It is like you totally missed the anti-globalism, wage inequality issue completely.
5
Oh YES ... I have been waiting for someone in the media
to take up this point ... the automation is a concern, like
so many other major problems we have today, but to
hand-wave at automation and then try to seduce the
country with a "Freedom Dividend" which is not nearly the
same as a UBI ( Universal Basic Income ) and try to dream
up all these other changes because one is good at math?
Thank God someone is trying to take a moderate realistic
stand on this subject.
4
Yes, Burnett’s was a bad question. But has even one political leader, of any stripe, at any level, given any thought to the cultural and political upheaval that will take place when cars drive themselves, and when stores have no employees (like the one at which I bought a sandwich tonight)? Automation will not destroy work in America. But if we don’t think about policies for an ever more automated economy, the disruption for Americans, especially lower-income Americans, will be awful.
14
My idea of an 'escapist fantasy' involves a cold drink on a hot beach somewhere. But I won't be able to afford to do it when my job gets automated.
3
I agree with Professor Krugman about almost everything, but he's partly wrong about automation. I work with software for self-driving vehicles; our business model calls for eliminating roughly 12 million jobs in American alone in the next 10 years. I'm referring to truck drivers, taxi drivers, bus drivers, Uber drivers, etc., who will be replaced by automation in the next 5 - 20 years.
He's right that automation is not, at this moment in time, the biggest driver of job losses. But it's likely that AGI (artificial general intelligence) will be capable of doing all jobs that humans currently do at some point. It's just a matter of when.
How can humans compete with machines that can work 24/7/365 without pay or rest? If those machines are capable of independent thought, creativity, and even emotions...then they will replace us all eventually.
Some people (who know far less about software development than yours truly) dismiss this future as a fantasy. It's worth remembering that many people of the past thought it was impossible to fly. When airplanes proved them wrong, many predicted that planes would remain a minor curiosity. Computers would never defeat humans at chess. And it was impossible to go faster than sound or walk on the moon.
75
@Thomas Hardy
Automation also create jobs. Robots do not drop from moon. Jobs like more teachers to educate engineers, scientist do design AI for robots, manufacturing and maintaining robots. But the question is: how many?
3
@Thomas Hardy
It seems to me that the basic argument here is
somehow made on statistics, that some AI driving
program over millions of trips will statistically work
out to be in some way safer than human drivers.
While that may seem convincing, the people who
make profits off this have had a way to snookering
us with fake-news and fake statistics for a long time
such that it has become business as usually.
If they can figure a way to force it on people as all
this other tech has been forced on us, maybe ... but
if I as a citizen have anything to say about, I don't
want to be driving my car aside a robot controlled
big-rig hauling a giant container of gasoline or
nuclear waste. No thank you.
7
@Thomas Hardy - just because your business model calls for automated cars, doesn't mean you will succeed any time soon. Your investors are taking a big risk with their money.
It is better to worry about actual problems today: excessive monopolistic power using outsourcing and technology to drive rising inequality. These are fixable problems.
6
I was with you until this sentence: "So harping on the dangers of automation, while it may sound tough-minded, is in practice a sort of escapist fantasy for centrists who don’t want to confront truly hard questions." Most likely the centrists are just mistaken, like Erin Burnett and the AP fact checkers. I'm not sure what good it does to alienate centrists.
3
I think this fails to account for part of Yang’s justification for UBI— data. Big tech has free access to everyone’s data and reaping a tremendous profit. UBI is a fair way to compensate everyone.
24
Agreed. Contemplating the future is well suited to science fiction but it's a diversion for politics. The focus should be on restoring American institutions of government and human dignity perverted by the seeking of excess profit by the continuous nibbling away at the individual quality of life.
3
@DAM So politics should not be about preparing for the future at all?
Also, Yang actually agrees with most of the goals you've listed. So please give Yang more consideration than a biased Krugman article or the brief but superficial coverage he usually gets from the press, including from this newspaper.
Thank you for bringing up that trade agreements have had an impact on US jobs. This was accepted for a while but now everyone talks automation. It seems these are both factors along with lower labor costs and less regulation. Although how does Germany stay a manufacturing powerhouse?
With all the trade agreements I don’t understand why we didn’t demand to bring our trading partners up to our labor standards? If we had insisted on a phased plan with NAFTA back in the early 1990s, both Mexican and US labor forces would have been better off by now, well maybe. The golden rule applies, he who has the gold makes the rules.
We really need some sort of ten year plan to force labor standards up around the world.
12
@spughie The US remains a manufacturing powerhouse. People misunderstand this. Manufacturing output has consistently increased while employment in manufacturing has consistently declined. The math is not hard here.
I agree, though, that it is both automation and trade that has decimated the manufacturing worker.
8
Dr. Krugman:
I'm not sure that "Economic history repeats itself incessantly to the centuries"...
Isn't it possible that the technological changes become qualitatively different from the old technological changes?
The XX century technological changes were gradual, provoking the move of millions of persons from the agricultural areas to the cities, without producing the horrendous excesses of the XIX century with ghettos, unemployment and hunger.
Maybe this time we could have an "involution" in that sense, and we are watching the disappearance of the middle class, the enrichment of the few and the increasing of inequality, poverty and hunger, including homelessness?
But now, of course, we have a strong government watching, with different tools to modify the course for good, though slowly and with a lot of pain: resistance from the right.
3
I do not see automation as dangerous, I do see the millions of truck drivers, Uber drivers, train engineers and others on our roads and rails without jobs in the next 15 years as dangerous. Automation is inevitable in a capitalist economy. What we do in response to automation is going to be key.
4
@SPH
Jobs will grow in services though one critical issue will remain. If they pay $7-10 an hours it will be a catastrophe. There is a growing need for higher valuation of labor.
2
While usually Krugman exudes brilliance, in this column he is living in a world of crusty denial.
11
Appeasing the Bernie Bros? They should be delighted.
Now, sadly, you're going to have to deal with Yang Gang blowback.
6
@republicon
The Yang Gang is far more turned on by his proposal of throwing money at you all than anything to do with automation. Who do you think will pay for all that free money each month. And Elizabeth Warren is criticized for being vague. Priceless.
If this were 1908, CNN would have asked, "what happens to all the people in the big cities employed to pick up after horses; all will lose their jobs because everyone will be driving this Model T thing, Is it even safe? What happens when one Model T hits another Model T."
And democrats would have answered: "yes, to correct that problem, I am going to: (a) stop Ford and ban Model T, (b) tax the wealthy, (c) forgive all people's home mortgage loans, and (c) outlaw firing of horse-doo collectors.
That is what I heard in the debate, just 2019 version.
2
Yang does a good job of looking like his heart is in the right place, and I’m sure his efforts to buy votes will go appreciated among Americans who have been cheated out of their basic wages for forty years. They need the money.
OTOH yes, his tech fantasies are ludicrous. I expect to see neither driverless cars nor trucks in my grandkids’ lifetimes, even. The only place the “4th industrial revolution” matters is in factory and warehouse automstion, and that’s already mostly happened. That will never change.
2
@Steven Given how often existing machines break down and how unwilling companies are to pay for repairs or new parts, I am also highly skeptical that we will all have self driving cars anytime soon.
1
Vonnegut warned about automation in 1952 and the years that followed were a golden age for workers -- so what? Vonnegut's work was science fiction, meaning he was warning about potential problems coming in the future, not in the 1950s. Welcome to the future, Paul.
Another point that Yang makes is that we should be focusing on quality of life. Maybe we should consider trying to structure society so that people don't have to work their butts off for most of their waking hours just to make ends meet. We have the technology. GDP is not the right yardstick to measure a society's success.
39
Krugman does not have a good track record predicting the future of technology. In the 1990s he said the Internet was no big deal. More recently he expressed surprise the wind and solar power have gotten much cheaper and are now competitive with fossil fuel.
Artificial intelligence and robots are fundamentally different from all previous innovations, because they compete directly with the last large advantage people have over machined. Now that bots can drive cars, they will soon learn to do all manual labor. Now that they can diagnose x-rays, they will soon learn to do most mental labor. In the past people migrated to new jobs, but sooner or later we will run out of jobs. No trend lasts forever.
See: "Humans need not apply"
https://www.youtube.com/watch?v=7Pq-S557XQU
9
@Jed Rothwell
It would have been much more useful for you to provide links that show Krugman actually said any of these things, and if he did, when.
Perhaps he did, or said something sort of like what you claim, but it is all too easy to put false words into someone else's mouth.
Right wingers do this all the time in the Krugman comments, posting false claims about Krugman saying or predicting X or Y, when I know what they are claiming is at variance with Krugman's established positions.
1
@Carey Sublette Krugman ruefully admitted saying both of these things in recent columns. I do not have an index of his columns, so I cannot easily find them.
He did not express opposition to wind and solar, but he said he was surprised at how quickly they fell in price. A person who understands the technology better would have known 20 years ago that the price would continue to fall. There was plenty of room for price reductions. There were no likely shortages of materials or suitable locations.
Since 2016 he has been an enthusiastic supporter of wind and solar.
Here is his quote about the Internet, from 1996: "The growth of the Internet will slow drastically, as the flaw in ‘Metcalfe’s law’ — which states that the number of potential connections in a network is proportional to the square of the number of participants — becomes apparent: most people have nothing to say to each other! By 2005 or so, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s.”
1
Given the untrustworthiness of this administration, its unqualified “acting” heads of vital departments, and its in-your-face corruption, its employment figures just don’t match the reality of 50 qualified applicants competing for one low-pay, no benefits job. Decent jobs (other than warehouse, trucking or the medical industry) just aren’t there
2
People love talking about robots like they can just come to life and take over the work of a dozen people. Here’s the harsh reality: the minute you bring a robot into your factory/warehouse/studio is the minute you have to add at least 3 people to your payroll. Programming a robot takes a fantastic amount of time, by people that are reeeeeally intelligent. Setting up the space and workflow to fixture the material under the robot in a repeatable way takes even longer. I’m a small business owner and I wish I could automate, but I can’t afford the labor capital that’s required. Robots are affordable. Everyone would automate if they could afford to hire the talent that’s required to run it. And no, I wouldn’t fire workers because of it, that’s not even an option. I would train them to operate software or handle the fantastic amount of work we’ll put out because of the new tool. And that’s the point: it’s just a tool. It’s like freaking out over the invention of the forklift because it replaced the need for ten dudes/dudettes to lift something.
2
Paul you were wrong about the predictable negative effects of globalization on working class jobs so I honestly don’t know whether to dismiss your unconcern about automation doing the same.
9
Well Paul, new algorithim designed to replace pundits has arrived, it must've written your column today. Mass job replacement is occurring right now,nand is accelerating quickly. Is your algorythim saying we should ignore and not prepare for the potential chaos that will ensue. History proves me more correct than you.
4
@Bill Gates
Is Microsoft accelerating the chaos?
Automation is real. Just not how they describe it.
So it looks like we are back to the shovel. And, I will always remember my World Civilization professor passing out a piece by Robert Hullihan about the invention of it and the ditch which followed - the inventor “sweating and gasping” as he demonstrates its capabilities, while the villagers wonder what this new invention will mean for them.
1
One small request to the esteemed Dr. Paul Krugman and the NYT:
Please create and place a pie chart on the front page for the next ten years showing the different causes and percentage of effects that led to our enormous income disparities since 1979.
Keep it simple. (America doesn't read any more)
1
"Yang has based his whole campaign on the premise that automation is destroying jobs en masse and that the answer is to give everyone a stipend — one that would fall far short of what decent jobs pay."
The ultimate problem Yang's proposals deal with is income/wealth inequality. Using value-added taxes along with universal basic income to help alleviate this inequality is not "an imaginary solution to an imaginary problem." In fact, the idea has support from another eminent economist who has also written about this in the Times.
https://www.nytimes.com/2019/09/27/business/yang-warren-taxes-mankiw.html
6
Conservatives say "I will give you tax cuts."
Progressives say "I will give you free things."
Andrew Wang dispenses with euphemism and says "I will give you money."
.
It's a ridiculous and cynical proposal on its face
@Bumpercar
What does it mean “free”? Is public education free?
“... the answer is to give everyone a stipend — one that would fall far short of what decent jobs pay. As far as I can tell, he’s offering an inadequate solution...”
As far as I understand Yang’s Freedom Dividend/UBI idea, that it “falls far short of what decent jobs pay” doesn’t make it an “inadequate solution” - that’s intentionally part of the solution.
The purpose isn’t to provide enough to replace a lost job, but a cushion. Or to paraphrase Warren Buffett (when asked how much money he’s going to leave to his kids), “Enough that they can do something, but not so much that they can do nothing.”
Yang’s UBI doesn’t replace a job, but gives you a little cushion to leave a job you don’t like, to take a risk and start your own business, or maybe to do something you really enjoy but doesn’t pay very much, like be a teacher or play guitar in the local brewpub. In addition, it gives a modicum of pay/recognition to all the stay-at-home parents, elder care givers, others working for no pay.
Krugman May be right about the tech revolution and the robots not coming for the jobs, and UBI might not be the best solution, but UBI isn’t a bad idea because it “falls short of what decent jobs pay” - that’s part of the plan, not a fault in it.
16
This article will be funny forever.
4
Capitalism will collapse if people can’t find work.
1
Oh there you go, bringing data into again. If only people would listen . . . .
Paul Krugman said, in 1998, “By 2005 or so, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s.”
Krugman was wrong then about the Internet, and he is wrong now about automation.
16
@Brett
Yes, he did say that and was clearly wrong - probably for the same reason he cites here -- he doesn't see the transformation showing up in the gross economic data yet.
But that is always a lagging indicator. Only when trends are very well established for some time do they begin to move gross productivity numbers. And once that starts its a game of catch-up to deal with the effects.
A clear example which we can, with high confidence project, is effect of automated vehicles. The most obvious target for these will be long-distance trucking since the trucks drive on highways, which are simple environments compared to in-city driving, and account for the majority of trucking labor costs. Teleoperators can bring the trucks to and from their end points for loading and delivery, and one such operator can handle many trucks in the course of a single shift.
Once long-distance trucking goes automated (and this transition will happen quickly once it begins) then all the support businesses to serve truckers - except for fueling depots - will go under as well.
5
Bravo. Well said, Professor Krugman.
Paul, there is an “nth industrial revolution” coming our way – if we could just get out of our own way...
Don’t believe me – or my commentist rants on nuclear energy over the past decade...
Your Governor and your Senator – obviously fervent readers of my fissile rhetoric, going way back – have come over to the greener side...
To say nothing of the subject of your lead-in pic...
Our federal government is actually beginning to move up the s-curve in this industry – to wit:
http://www.world-nuclear-news.org/Articles/DOE-NRC-collaborate-on-advanced-reactor-deploymen
PS
Though much of the PPPL equipage would make splendid coral reef, the supercomputing advances that’re finally happening there and elsewhere – to wit:
https://energynews.us/2019/10/16/midwest/how-supercomputers-could-help-crack-the-code-for-advanced-nuclear-reactors/
Should make SMRs happen on a schedule (at least) 3X faster than is currently being admitted...
And yes, we could give the Chinese a real run for their (our?) money, in this space...
This will be a $1T industry, within 10 years...
With many times the number of heavy-industrial shipyard-like jobs we now have in the US...
LHC didn’t do much for the particle industry – but it gave the EU a reason to build really strong magnets...
Which are fundamental to higher MRI resolution, among other things...
Along that line, several other uses – some already transformational – for the plasma physics know-how...
2
The workforce that is the most vulnerable is made up of lower educated and often innumerate people. The arrival of the undocumented people looking for work will mean that the lower end of the work force will be increasingly lacking in skills suited to manufacturing. It seems safe to say that there are still job openings in the gardening field and that these openings will continue to be filled with the undocumented workers without social security numbers and vulnerable to just about every downside the society has to offer.
1
It's too bad that Paul Krugman has conflated Andrew Yang's two big ideas. As for the first idea, I agree with Krugman that Yang's doom and gloom about technology resulting in net job loss is alarmist and historically short-sighted: human beings have infinite ability to conjure up new ways of deploying human labor. Since the beginning of human civilization, our labor market has been on a constant churn of creative destruction and generation.
As for the second idea, I know with absolute conviction that the percentage of Americans who cannot be part of the labor market is significant and growing. People who are addicted to substances, mentally disturbed, suffer from chronic health issues (including obesity), and have myriads of conditions that simply make them unable and unwilling to work make up 22 percent of Americans who are not working even when the country is experiencing historically low unemployment rate. For these people, some universal income program is the only way for us to give them the purchasing power to spare them the horror of living in squalor or on the streets. For anyone with their eyes open, American cities and small towns are filled with broken Americans who are not broken enough to be in an institution but too broken to work. The government has shown that it's incapable of any meaningful intervention for these Americans. The ONLY way to give them any visibility and leverage is to empower them with money. $1000 a month will make them matter.
3
Manufacturing lends itself to productivity increases, our service heavy economy does not. Maybe compensation shouldn't be based entirely on productivity. Salaries for some service industries have risen faster than improvement in productivity others have stagnated. The difference service. teachers and healthcare Unions. As long as we live in a country where Unions are demonized and frustrated by nonsensical anti-union laws you will see stagnant wages. Think of it as another form of Trickle Down Economics. economics that makes sure wealth trickles downs This one might actually work.
2
When I took my first job in Library Tech Services 50+ years ago I spent most of my time ordering, sorting and filing catalog cards which was boring, error prone and wasteful. When we moved to our first computerized catalog we were able to shift our efforts to helping patrons find the information that they wanted and needed. Over the next 50 years the ever more capable librarians and library information systems made vast and readily observable improvements in the quality of services available. Automation is not the enemy.
5
I think Professor Krugman should pull himself down from the macro-economic stratosphere, and try to relate to people as individuals... as do the politicians he critiques, and not as components of statistical abstractions.
The fruits of automation are already among us: as evidenced by the inequality of wealth, the decline of the middle class, and the inability of most non-NYT readers to afford college bound children, without the need to incur massive debt.
4
Resist! Don’t use self checkout. Put your check in the envelope, stamp it, and mail it back to pay your bill. There used to be self-driving technology— a well-trained horse would get you to shelter if you became incapacitated.
Faster is not always better. Sometimes it can even be worse.
2
@Lawyermom I agree in spirit. But every time I go grocery shopping I look over at the checkout line and over half the lines aren't even open. There's a handful of people in a few lines, often with full carts. And I'm a single guy with a small basket who just needs to get in and out. "Don't use self checkout" isn't a realistic option.
As for put your check in the mail, autopay is a lot more reliable and requires next to no effort. I resist bad automation and social media. But if a process has been automated in a way that is substantially more efficient, I use it. Keep in mind, I got a busy life and career where I got to compete with everyone else. I don't have time to waste an hour a month on bills like my parents did.
As for self-driving cars, I'm never giving up my human-driven Honda Civic.
2
@Lawyermom We don't have to practice ludditian rituals to keep people employed. There is near endless work in front of use converting mankind away from the use of fossil fuel.
@James I guess your time is more important than the jobs of supermarket cashiers and postal employees. You are proving Andrew Yang’s point.
Software is continuing its course to eat the world. Future job losses due to automation will impact more professional white collar jobs than blue collar.
3
@Big Lebowski
thanks for the info dude.
1
"Burnett declared that a recent study shows that “about a quarter of U.S. jobs could be lost to automation in just the next 10 years.” What the study actually says is less alarming: It finds that a quarter of U.S. jobs will face “high exposure to automation over the next several decades.”
That's a distinction you may be comfortable with, but a family struggling to make ends meet month to month don't have the luxury you have from where you sit.
And in the late 1940s few American jobs were being shipped overseas, so those out of work had the ability to move to new employment. Now that option is essentially nonexistent, as any new job markets or technology are going to be as automated as possible from now until eternity, or until people get fed up with it.
So yes it is different. And scary.
3
Yang has made the automation scarecrow such a central tenet of his political philosophy that he seems unable or unwilling to envision a wider agenda for the country. This is why he has such a narrow base of support and why me and many others cannot see myself supporting him. His constant use of racial stereotypes doesn’t help, either.
1
We were taught in Social Studies class in 1960 that because automation was improving productivity it would allow the breadwinner to work fewer hours per week and have more leisure time to spend with his family.
Instead we now have a society in which the BENEFITS OF PRODUCTIVITY INCREASES have NOT TRICKLED DOWN into the workers pockets through wage increases and more free time but HAVE BEEN VACUUMED UP by the CEOs and stockholders.
97
@Semi-retired Capitalism ensures those productivity gains go only to the owners and not to employees. Instead, workers lose their jobs to tech instead enjoying the results of their work.
This is an area Marx had right. More worker control would actually lead to a shorter work week and more leisure time. Capitalism will never get us there.
19
Just an observation but it seems people are demanding many more services no longer wanting to gather their own groceries. They subordinate other shopping duties to online pick up and delivery companies. Many don’t even know how to make good coffee and purchase a daily cups of brew. Seems all of those businesses are hiring.
2
@ Thomas Jordan: But are the jobs well-paying jobs with benefits? Usually, no.
Automation gives one person the power to do the jobs of many. That makes that one person more valuable to employers, but (as we have seen) that doesn't mean that employers will pay him/her more.
In fact, because fewer workers are needed, the market price (wages) of labor declines. It's a supply and demand thing--greater supply coupled with lower demand means lower wages.
At some point, human labor may not be required at all, diminishing its value to near zero (it will still have value so long as its cost is less than the machines that replace it).
Economists assume that any resource that exists will be used, at least so long as its contribution to production exceeds the cost of 'extraction.' But humans have to eat--and in order to be effective workers, they require education, etc. So there are certain minimum costs required for human labor extraction.
Also, being able to produce more also does not mean that we will be able to consume more. While much of the world is needy, overproduction is a problem. (Partly, but not completely, due to distribution problems.) So produce though they might, surplus workers may not be able to sell the fruits of their production.
But this is also the area where automation will ultimately help the rank and file: over production should lead to lower prices, freeing us, to some extent, from the demands of work. The problem is that we have allowed monopolies and oligopolies to flourish, artificially increasing prices.
5
I always learn from reading Paul Krugma's columns, but this time I'm not sure I agree here. From the perspective of recent history, it doesn't seem right. Jobs are indeed being automated away and this isn't going to stop. Plus, I like Andrew Yang - a lot. He speaks clearly and directly and has an understanding of the way technology is changing our lives and the world as does no one else on that stage. However, I agree that we should not go down the rabbit hole.
3
There’s a lot to agree with here, but there seems to be a disagreement that’s getting glossed over. Yang’s proposal and Krugman’s proposal are different kinds of the same thing: public spending. This is a disagreement over which kind would help more.
Traditional public spending and a universal basic income (UBI), as alternatives, have different pros and cons.
Paying for them, as with Medicare for All, is not necessarily a matter of raising taxes. As I recall, another column on MfA found enough money in defense spending to give every American ~ $2,800 per year.
The UBI would write them a check. Traditional public investment would order an infrastructure project, grant or loan some amount based on public and/or private bonds, and pay owners a large cut that would trickle down to their employees and contractors.
If the point is stimulus, cutting a check would seem more potent, vs some portion going to a subset of workers holding signs or slinging jack hammers, mixing cement, etc. It might be more equitable too.
If the goal is return on investment, however, a well targeted grant could net a better return for the US economy than the UBI, which might get invested, saved, spent locally, or on imports that largely help other economies.
Perhaps tariffs could offset this UBI loss to an extent, but finding ways to encourage local circulation for this investment could be very important to seeing it lift all US ships.
I hope Mr. Krugman addresses this disagreement in the future.
7
Going forward, everybody is looking for Eternal Truths but nobody can quite agree what they are. A few candidates from out of left field:
1) Too many people now live on the planet to sustain its resource base at current levels of consumption.
2) With increased technological innovation, including automation and AI, ever fewer people will be required to produce the necessities for human existence.
3) Maintaining stable global markets at or near full employment requires ever increasing consumption of unnecessary goods and services, which in turn requires incessant artificial stimulation of desires for inessential pleasures.
4) Some of the unnecessary goods and services required to sustain a global market equilibrium are actually harmful to humans, both physically and mentally -- a fact that we are now experiencing at exponentially increasing rates.
If the foregoing is accurate, the basic issues are ones of rationalizing the fair distribution of essential goods and services, removing profitability from the production of harmful inessentials, and inventing meaningful new forms of work that are ecologically neutral in their impacts.
Perhaps the answers will bring us full circle, to the point of creating a democratic and inclusive cohort of the "philosopher kings" envisioned by Plato. What machines can never provide are an ethical society and a feeling for the creative relationship between truth and beauty. Only humans can do that.
This challenge is also our opportunity.
14
I recently went to a small workshop for academics to discuss the issue of automation and labor and more than one presenter highlighted this very issue. Great to see you bringing it up as well!
1
This is obviously a complex problem. And with such problems, there is rarely a simple answer. Krugman spends much time in this column estimating the size of each component of the problem: automation, trade policies, jobs moving overseas, etc. The real solution is determining which components need to be included and blended into a workable policy. One overlooked component has been to drive demand for U.S. and locally-made products. We need to educate the public on the true value of buying a domestically-made product, including all the economic and social benefits. There are nascent buy-local movements in America now. Unfortunately, Americans over the past decades stupidly bought into the premise that a cheap hammer made in China and sold in Walmart is good for our economy (this is also espoused in may economics textbooks). But we all know it was far better for China's economy, because the lion's share of that sale went overseas. The rest went into Walton bank accounts, with a tiny sliver feeding local economies. There are other components to this problem as well. Let's approach this logically and holistically and not in a dogmatic like Mr. Krugman.
3
"the persistent weakness... isn’t about automation; it’s about inadequate private spending." No surprise; when so many people are struggling to make ends meet, they buy necessities. 75% of our economy is consumer spending, which is constricted when so much of the nation's income goes to a small percentage of the population.
25
Acemoglu and Restrepo disagree with Krugman and, quite frankly, seem to be more research active, and less politically partisan, over the past two decades.
In particluar, note their conclusion that the negative effects from automation "indicate a very limited set of offsetting employment increases in other industries and occupations."
5
So well said! And, this guaranteed income of some minor amount per month for each person never made any sense to me for a variety of reasons, including the fact that if people had more pocket change, prices would likely increase and it would have little, to no impact, except wasting resources that could have been put to good use.
2
Keugman is right about the data and the bad question.
The tools are in hand to adjust trade relations, to improve taxation efficiency and close loopholes which can provide the government spending to tackle infrastructure spending that can lead to new jobs across the country.
It’s also fair to admit that some types of low skill low wage jobs are more exposed to the risks of automation than others. Trucking could be one which takes a big hit disrupting or reducing the wages of legions of drivers.
1
I’ve always thought Yang’s idea was mostly a single policy and not something to choose a candidate based on. The context of history made here makes it even less than a policy and essentially wipes away any reason to consider Yang. Every question that is asked of him comes back to the same answer from him - $1000/month. People want jobs - not handouts. Jobs give people dignity and purpose. Let’s stop talking about this and move on to substantive issues.
9
@BobB Have you actually looked at his platform, though? $1000/month weaves a pretty good solution for multiple problems, all of which are related. And, he has a ton of other very good ideas besides this. Yang deserves more than a cursory evaluation.
Also, the technological change is absolutely coming and will batter us. Krugman of all people should know that "past performance is not an indicator for future performance." The technological change we speak of is an intelligence that can and already has matched and exceeded human capability. It's not a surprise most people who support his vision are techies and young people.
The best thing about Yang's candidacy, if you look beyond the surface, is that he asks the fundamental questions. What is value? What is work? If you actually take time to listen to some long form interviews (or his AMA tomorrow), maybe you'll be enlightened.
3
@BobB
People aren't going to get good-paying stable jobs with long-term security anymore. That's the whole point. UBI is a safeguard against having to relocate 5-10 times throughout one's career, navigating the less and less reliable corporate "health care" system, and stagnant wages.
Your quick willingness to disregard an extra $1000/month bespeaks your economic security against a phenomenon that may be posed to wipe thousands of families into joblessness, poverty, addiction, and mental health issues, the cost of which will have to be shouldered by...guess who..the taxpayer. I'd rather give everyone $1000/month and let them have a fighting chance to keep some of their dignity.
5
Back in the 1970s, it took something like 100 hours of human labor to build a car, now it is closer to 10. The difference isn't just robots, but changes all along the supply chain to reduce unproductive tasks, especially moving material around, and in-process defect reduction, reducing wasteful scrap and rework at the end of the final assembly line.
Industrial capacity utilization at this stage of the economic cycle used to approach 90%. Now at full employment, it is at 77%, again because process improvements have increased output from existing industrial facilities.
Manufacturing has been the innovator in process improvement and defect reduction but every other field, from health care to construction to education will face similar pressures.
If you have ever watched a building been built, you know what I mean. Most houses in the future will be prefabricated in modules in a factory instead of built from scratch on-site.
Financial services and retail are already undergoing huge changes.
There will be new jobs, especially in design, repair, maintenance and upgrading of technology. But employment is going to look vastly different in the future. Just ask a Millennial. They can explain it better than us older folks.
8
I think that there are a lot of things that robots can do and a lot that they might not do so well. But the old idea that robots might function to provide leisure for people becomes risible when people whose jobs have been taken away by robots can't afford robots because they no longer have paychecks. I understand that most robotic job taking takes place in the manufacturing arena but it also takes place at your local supermarket where the cashier is replaced by computerized checkout. I honestly believe that people need to work and not just for a paycheck but to put value into their lives. And, I honestly believe that work should be at least somewhat fulfilling, that somewhere in the work day or week there is a moment when you can say 'yeah, I did that, I made that." This country needs a huge commitment to infrastructure whether it be roads and bridges and tunnels or a major restructuring of mass transit and rethinking of the energy and climate cost to the world of the single commuter in their car going to the grocery store for a quart of milk. I'm guilty of this everyday. Mea culpa.
9
The robots haven't captured all of manufacturing, and I definitely have my doubts about self-driving cars, but automation continues to destroy other types of jobs.
Secretaries and other clerks are mostly gone. There are way fewer bank tellers than there used to be, because nearly everyone banks online. The number of people working at the Post Office has dropped by hundreds of thousands over the last twenty years. And when was the last time you went to a record store?
10
Paul is consistently wrong on this. Capital takes precedence over labor in our world. If you have the bucks you either build robots, or you set up shop in China or Malaysia. Our workers lose either way. There is no stopping automation, and there is no stopping globalization. If you don't have the education to either build or repair the robots you are stuck at the Wendy's counter. And that job is leaving soon too. UBI is a means of cushioning the fallout of a rapidly changing economy, and it is not meant to replace low paying jobs. It is a forward looking, if misunderstood concept; our good professor is stuck in the past.
93
Paul, I am a retired sheet metal worker. When I started in the trade there were 6 men in the shop cutting duct, they have now been replaced by a plasma cutter, one Man. The auto manufacturers, the car makers, produce a car with ninety percent less labor than 10 years ago, the coal miners mine a ton of coal with 90% less labor than 10 years ago. Automation is a problem and it is increasing at a exponential speed. Amazon has fulfillment centers that mostly employ robots. The docks no longer use longshoreman to offload the containers, computerized trucks shuttle the containers around, no men. Algorithms will soon replace engineering jobs. Wake up, go out and see a factory.
267
@Bill T - you say, "Algorithms will soon replace engineering jobs.". But your examples are of manual labor, not of engineering jobs. Engineers are in higher demand now than they were in the last decades.
13
@Mark Lai
Not all engineers.
I remember in the 60s speaking to an architect who told me a computer will never replace the engineers and draftsmen in his 300 man office.
Ummmm, twenty years later, computer aided design (CAD) reduced his office to 150 workers.
The engineers that are in high demand are not the engineers who got their degrees thirty years ago. The same applies to IT workers laid off at age 55 who are no longer up to date.
10
@Mark Lai My son with a PHD is a nuclear engineer and scientist working at Los Alamos national laboratories. He was tasked with a project crunching dada that had been compiled over the last 20 years, it was expected to take six months. He created an algorithm and it was done in three weeks. The older and more seasoned engineers looked on in amazement at the efficiency. No matter how you spin it this is a loss of jobs.
9
Dr. Krugman, I think the discussion is worth having, not so much because we are certain that new AI based automation will take away jobs faster than new jobs are created, but to prepare in case it really happens. The potential over the short term may be a bit over-hyped, but over the long term, it's real. And I think you would agree that job elimination due to automation vs. job creation due to innovation aren't automatically in some sort of balance.
So the useful discussion is, who deserves a share in the enormous wealth that could be created by a fourth industrial revolution based on machine learning? Is this a situation akin to the Alaskan oil dividend, where the people who live on the land are entitled to a share in the wealth? If we're all members of a society that educates its children to work in the field, provides infrastructure for the creation of new business, and encourages innovation, do we deserve anything in return for all our investments? Or do all the profits go to the capitalists who put down the first stakes?
I'm not planning to vote for Andrew Yang (although if he can breathe and he ends up running against Trump, I will), but I don't think his point is so easily dismissed. What does the future look like if we can't come to some sort of social pact in which we agree that we're all in this together?
27
Thank you, Dr. Krugman!
Personally, I tend to think of the projected Automation of Everything the same way I think about our nuclear weapons capacity: just because you have the ability to do something doesn't mean it's what you should do!
1
Mr. Krugman is partially correct. The trend now and for the next many years is not that people will necessarily lose their jobs to automation. Rather, workers will need to move "up the food chain" of a system that includes automation. Machines won't do everything, but they'll do part of what we are doing now, and humans need to retrain and up-skill to avoid the parts of the system that are in fact being cannibalized by automation. We can't ignore it, but it's not clear that we are doomed to be automated out of jobs ... yet.
4
Sorry Paul but I have to disagree. Although worries of past centuries about the employment impact of the Industrial Revolution, and later the advent of the automatic computer were misplaced, we now are facing the fact that even jobs that have heretofore required humans are going to go away fast. As a businessman I can tell you that if we assume a worker costs a total of just $40,000 a year, I would gladly pay $400,000 to buy the technology that would do that job. Lots of business people feel the same way.
13
@Thomas Smith
Why would you pay ten times more for technology to replace labor?
just curious
Because the salaried worker costs more in the long run: multiply $40,000 per year times X number of years. At 10 years, that worker costs as much as the robot, but produces less (isn’t indefatigable: needs vacation, lunch breaks, sick time, etc.). The trend is and has been replacing the most expensive cost of production—payroll—and technology is and has been shrinking payroll. Think how many fewer cashiers, bank tellers, travel agents, etc. Now think, if Expedia can book my plane tickets, and the Self Checkout replaces cashiers (so only one human is left to “mind the machines” instead of ten humans doing the jobs machines do now)... if that’s already happened, what’s stopping technology from increasingly replacing higher-order jobs like truck driving. Capital (owners of industry) wants reduced labor costs, because labor is the biggest expense in production. Much of the innovation involves the consumer doing the job that tellers/cashiers/etc. used to do for the consumer. If you look at it that way, then UBI is actually remuneration for us having to do our own self-service everywhere we go to shop, eat, etc. And furthermore, when the consumer is the product, as with consumer data being harvested and sold, shouldn’t we be paid for our role in creating value and wealth for those companies (Facebook, Google, and every other major player in digital commerce). Like Social Security or Medicare, which we pay into and then get paid out of in retirement, UBI does the same but here/now.
2
Not sure what reality Mr. Krugman lives in. I'm noticing more and more jobs being replaced by automated kiosks at banks, movie theaters, and of course the supermarket. My local bank branch in downtown Brooklyn two years ago had up to 4 human tellers working at the same time. When I went in for a deposit a few days ago, large sophisticated Automatic Teller Machines had replaced the human beings. Same for a local cineplex. I doubt that any level headed person is worried about actual robots taking over. No. We are more worried about the loss of jobs and dignity afforded to the working middle class. Nevertheless the increasingly escalating sense of alienation upon bumping into a robot aisle cleaner at the local Stop n Shop.
71
@M
Sorry, but the facts don't back it up. The number of bank tellers has actually increased since ATMs were introduced, even when adjusted for population growth. Ditto self-checkout lanes and cashiers. Even if those machines did displace workers, they eliminate low-wage workers but rely on skilled, higher wage workers to install, operate, and maintain. We might not live in a mass production economy anymore, but many of our industries - entertainment, food, retail - still rely on mass consumption that cannot be sustained by the 1%.
We've had this great automation debate before and the resulting Presidential Commission reported that technology eliminates jobs, not work.
6
The problem is the relation of production to income. Generally we think people must work and sell their effort or the results of their work in a market. But what if we took GWB's notion of the ownership society seriously? How can we link consumers to an income without doing work?
Well, that is what heirs to large fortunes do. And yet the 85% of working people not in management have not had a real raise since the mid-70s, even while productivity has climbed steadily. What really prevents all that wealth from accruing to those whose work created it?
Of course I appreciate the moral corruption the working class will suffer when their do not have to work any longer for their income, but that is a price we will have to see if we can bear up under.
6
I respect Mr. Krugman, but in this case I disagree. From his secure perch, Mr. Krugman seems blind to huge changes that are afoot in the lives of working people. I doubt he has had to look for a job recently; the experience is frightening. And the experience of having a job has also deteriorated markedly. Aggregate statistics like productivity hide a vast amount of underlying detail.
Mr. Krugman accuses Yang of fantasy, but the burden of proof is on him to show that AI and automation are not going to cause structural changes in the economy, and his arguments fail to convince. Self-driving trucks alone, which are already used in a few places, could cause huge disruptions if Yang is correct that truck driving is the most common occupation in 22 states. To date, the people who have been displaced by automation tended to be at the lower end of the socioeconomic scale. In the current economy, they have few, and bad, options to survive when they lose a job. An increasing number of us will be in that situation as the influence of AI spreads into white-collar jobs (I can clearly see it coming in my own profession). The "dividend" proposed by Yang is a humane and reasonable way to combat the displacement which avoids the burden and horror of trying to meet people's needs with a vast complex of government agencies and programs.
177
@Gkhan
AI will certainly cause changes in the job market. But the real point is technological change has always caused disruptions-witness buggy-whip factories and stable workers...but the overall number of people employed in transportation is still substantial.
While the UBI is a great idea, K's bigger point is that job weakness is related more to inequality and lack of salary growth despite productivity growth. People would rather have a good job than $1000/mo income.
For employees and the nation, bigger is not better. Monopsony employment, for example in security guards, and monopoly production makes the labor and goods markets less competitive and less functional.
Public policy should oppose consolidation, not through the ineffective and long delayed actions of antitrust law, but by shaping the market. A graduated tax on GROSS corporate income would do the job-despite $12 BILLION in profits Amazon pays no taxes, while they destroy smaller businesses daily. A 20-30% excise tax on the GROSS income of huge corporations would make local businesses-who also employ more people and pay higher wages-competitive again.
10
@Gkhan - you say the burden of proof is on Krugman to "show that AI and automation are not going to cause structural changes in the economy". But Krugman is not saying that. He is saying that the persistent weakness in the economy is not because of automation but because of inadequate private spending. He is probably right.
9
Thing is large scale public investment merely treats the symptom. The actual problem remains.
We still don't have AI and when/if it comes it will take time to have much impact. Every truck on the road won't be replaced immediately a driverless version is available.
And if everything is replaced by AI & Robots and people have no work or income how will the AI & Robots be funded. If you can't sell what they produce why have them?
2
There are so many gushing accounts of a robotic future coming out of that fantasy land called Silicon Valley that I don't blame people if they fall for all the hype. Just take self-driving cars, one of the biggest hype jobs in recent memory. Supposedly there were almost here. Now people in the auto industry who fell for the hype have run smack into reality and it will be at least decades before these vehicles arrive if they ever do. Of course there is no real need for them. People make excellent drivers and do not need to be replaced. Robots are very poor at many tasks that humans do easily. It is impossible to see the future but right now people should not get worked up about robots replacing workers on a mass scale.
2
I'm one of those weird Deomcrats who would be equally happy with Bernie, Warren or Yang getting the nomination.
A UBI is a good idea even without the looming threat of automation breathing down the neck of workers in jobs based around repetitive manual (driving) or cognitive (finance and accounting) tasks. Looking across the spectrum of policy ideas on offer this cycle - and there are a lot - nothing would put resources in the hands of the working class more efficiently than Yang's UBI. I love Bernie but I think Yang won the UBI versus FJG argument hands down.
Also - fellow Democrats should look at how well Yang is doing appealing to Trump voters in the Rust Belt, i.e. the Upper MidWest, i.e. the place Democrats absolutely need to win.
76
Exactly.
1
Seems most people who have an opinion on this have never tried to staff a factory making things. I have and I can tell you not a lot of people want to actually do these jobs. We have always paid at least $3.00 an hour more than minimum wage to entry level non skilled workers and well above the median for more skilled people. Often the job has gone unfilled leaving existing workers to pick up the slack or quite often it is a revolving door where we hire someone only to have them move on in a number of months or maybe a year or so. Offers to stay in the form of money or flexibility do not work and we have a very easy going environment to start with.
Automation has been a godsend to us. We are able to produce what we need to keep our customers happy and quality and profits are way up. It is still basically a nightmare to hire the few people we still need and I myself as a 61 year old proprietor work way harder and longer hours than I could possibly want to.
The funny thing is though we deal with a large amount of support businesses employing way more people than we would ever employ. Tooling suppliers, repair people [robots may not complain but they can get quite sick], programmers when we get in over our heads and the machines we own are made by a company in California with hundreds of employees.
So while we may be a small mom and pop manufacturer the automation we use creates thousands of jobs.
36
",,, has lately been growing much more slowly than in the past; it rose less than half as much from 2007 to 2018 as it did over the previous 11 years." Think the GFC may be a factor in that Professor? Oh, and how do those previous 11 years compare to the last hundred years? Manufacturing, resources and farming, the Republican job base, are objectively losing jobs.
1
Redistribution of wealth from those who have vastly more than they need to those who live from paycheck to paycheck -- that makes sense. But is an economic system that depends on ever increasing consumption viable in a time of climate crisis and environmental collapse? I get a queasy feeling whenever I hear about "insufficient demamd".
Our capitalist system's success has long been subsidized by easy access to fossil fuels and by plundering the environment. One way or another that will end. Can capitalism survive without those resources? Can we?
As to the effects of automation and AI, I can only say I hope our robot overlords get here soon, because we humans aren't capable of managing the planet we live on.
6