Few topics spike the ire of American voters like jobs, immigration, and trade, no doubt because the three are inexorably tied together, at least in the rhetoric of politicians who point to immigration and trade as the villains of the American jobs story.
This narrative was clearly in evidence during the presidential campaign, as candidates in nearly every race made admirable pledges to “bring back jobs” through any number of means: reducing immigration by building a wall on our “southern border”, tearing up trade deals, punishing companies that relocate jobs outside of the US, and/or rounding up and deporting undocumented workers. That political rhetoric has spilled into the post-election zeitgeist in the comment sections of online media, like our Immigration, We Simply Cannot Afford This post that examined the economic contribution of legal immigration over the last half century. The comments section was a river of outrage over illegal immigration and trade, even though neither topic was the focus of the piece. Still, the post hit a cord and has reached over 500,000 views.
It’s no wonder. Jobs — the lack of them, generally, and the lack of good-paying ones, particularly — is a hot button for many Americans who are suffering from stagnant wages or worse, up-ended careers. While immigrants and trade may be convenient whipping boys (why else would politicians lite on them so readily?), the preponderance of evidence suggests that it is automation, not immigration, that is eating American jobs.
Jobs Lost To Bad Trade Deals? WRONG!
The idea that we can “bring back jobs” by renegotiating trade deals to restore lost manufacturing jobs is simply wrong. Still, the concept is appealing. It evokes images of an anthropomorphized state sneaking into the United States under cover of night, packing crates full of jobs onto cargo planes, and shipping them off to some unfamiliar land where they will toil in darkness until the bold politician sends a rescue mission to liberate and “bring them home.” It makes for stirring campaign rhetoric, particularly to the un- and under-employed voter. Unfortunately, it is a tug of war with the past. Some of those mostly vulnerable, sadly, understand it the least.
We focus initially here on examples from the manufacturing sector as it is the first segment consumed by automation. As Ben Casselman of Nate Silver’s group FiveThirtyEight summed it up, Manufacturing Jobs Are Never Coming Back.
88% of manufacturing jobs were lost to advancements in technology and automation (Ball State Study)
Automation and advanced technology – more than any other factor – are responsible for job loss across the United States. That’s the finding of Ball State University’s Center for Business and Economic Research’s report, The Myth and The Reality of Manufacturing in America, which reported that 88% of manufacturing jobs were lost to advancements in technology and automation. That’s hundreds of thousands of jobs moved from human to machine labor. The American Enterprise Institute clearly states the situation in output and employment figures: In inflation-adjusted constant 2014 dollars, US manufacturing output has increased more than five-fold over the last 67 years, from $410 billion in 1947 to a record-setting level of output last year of $2.09 trillion. From a peak of nearly 19.5 million US factory workers in 1979, the number of manufacturing employees has steadily declined to a recent low in 2010 of 11.6 million workers before rebounding to slightly more than 12 million employees last year (2014). This is not unlike our transformation from an agricultural society to and industry society, AEI notes: With fewer than 2% of America’s workers, we produce more agricultural output today in the US than when much greater numbers and much higher shares of the nation’s employees were working on farms. Jobs haven’t been outsourced; they’ve become outdated.
Just as farm workers moved to factories, so too did manufacturing workers shift to find replacement work in the available space of retail. Unlike the move from agricultural to industrial, the impact was a significant income loss from high or middle wage to low wage jobs, and there is understandable frustration. The next wave of automation is now beginning to impact the retail and service sectors as evidenced by Wendy’s restaurants recent announcement to replace counter service workers in their stores with kiosks. Wendy is testing this concept, initially, with 1000 stores or 16% of their stores. Bill Gates, Steven Hawking, and Elon Musks warn about the velocity of this change. Boston Dynamics is inventing robots that may replace humans more swiftly than we are prepared to respond.
True, work used to be simpler. When we ours was a local or national economy plodding along a linear productivity curve, jobs often did leave the country as U.S. firms sought out lower-cost providers. Today, though, we live in a hyper-connected global society where jobs are lost to intersecting and amplifying trends: atomization, augmentation, and automation. These trends have been accelerated by a corporate shift that began in the 1970s to prioritized shareholder returns (profit) over customers, employees, and community benefit (value). This shift alone is the key contributing factor to the greatest level of income inequality in the US since the 1920s and why, on both the left and right, voters were attracted to candidates who spoke of movements and revolutions. It is also no wonder that Bill Gates recently proposed taxing robots and Elon Musk suggests we may need to institute Universal Basic Income. Both those thought leaders can see where rapidly-moving automation is taking us.
Where Did My Job Go? Atomization and Automation Took Them
Today, the tasks of many jobs – particularly those at an entry level, but increasingly those in the professions – can be broken into separate, discrete pieces. This is the atomization of work. Once a job has been atomized and the routine and predictable components digitized, the atomic parts of a job can be parceled out to a global workforce willing to complete a task at the lowest cost.
As those tasks become more certain and their outcomes clearly defined, they will be augmented in whole or in part by computerized labor. The robots that supported and often supplanted factory workers (the focus of the Ball State study) have now moved into the professional ranks, where computer algorithms analyze reams of data far faster than human workers. Most certainly, virtually every job will be computer aided in the not so distant future. This is the augmentation of work and it will continue on this course until, ultimately, most tasks are captured in algorithms that execute jobs faster, more predictably, and more efficiently than even the lowest-cost human worker. Algorithms augment until they replace many human tasks and skills. This is how atomization and augmentation of work are interconnecting and amplifying the transformation of work. Those jobs most in the crosshairs of this phenomenon are those jobs based in either routine cognitive or routine manual work.
While technology is subsuming many job functions, this isn’t all bad news. In fact, there’s a lot about which to be optimistic. Aided by automation, we are producing goods and services at lower and lower costs making even advanced products more affordable to more of the populations here and abroad. More importantly, rising machine intelligence advances our potential to solve the most complex and threatening problems, from disease to climate change to the efficient use of resources. Working in tandem with technology, people can create, solve, and shape the world for the benefit of all. The process of reaching our collective potential will require a new approach to education that is absolutely lifelong. Research by GSV (Global Silicon Valley) finds that education expenditures as a percentage of GDP may rise from 9% to 12% in the next decade.
In fact, we will need both advancing machine capabilities, a more educated workforce, and immigration to offset population declines and care for our aging society in the developing world as our demographics shift dramatically, particularly in the U.S.
What Is the Purpose of a Company?
Over the last quarter century, the purpose of the company has changed from one that aggregated work effort in order to optimize productivity and create value for customers to one that aggregates profitability in order to create value for shareholders. The concept, known as the Corporate Profit Imperative, is now taking some heat. Indeed, now-retired GE Chairman Jack Welch finally called it “the dumbest idea in the world,” as Steve Denning reported in Forbes. The corporate profit imperative shifted the workforce from an asset to develop to a cost to contain as companies created more and more financial value with fewer and fewer humans.
MIT economist Andrew McAfee draws an even starker picture, which he refers to as the Great Decoupling of the US Economy. Using Census Bureau and Bureau of Labor Statistics data, he maps the diverging paths of automation and real employment. Real productivity rises with the helping hand of automated systems. Meanwhile, the net number of jobs and median incomes were stagnant over the last 25 years.
While not the only contributing factor, this fundamental shift helped quadrupled the market value of U.S. companies since the early 70s, as measured by growth in the Dow Jones Industrial Average from 1990 to 2015. Across that same period, the U.S. labor force has been virtually stagnant. Further, according to Pew Research, in recent years we have lost 6% of the middle class and added 5% to the highest income bracket and 2% to the poorest bracket.
94% of net job growth in the past decade was in the alternative work category,”(Lawrence Krueger, Princeton)
What About the Longest Streak of Job Growth On Record?
The last administration created the longest run of consecutive job growth on record. But, according to a report co-authored by Lawrence Katz and Alan Krueger, researchers at Harvard and Princeton, respectively, nearly all the 10 million jobs created between 2005 and 2015 were not traditional nine-to-five employment. We are moving toward a line in which more work is done outside the auspices of traditional employment wherein a company provides a consistant salary, health benefits, disability insurance, and other benefits traditionally tied to a company job. There are upsides to flexible and on-demand work afforded by atomization and first-generation work platforms like Uber, Upwork, TaskRabbit, HourlyNerd, and the like. These platforms enable individuals to work on their own terms and open opportunities for rich, global collaborations. They have tremendous potential to provide meaningful work and learning, and to re-imagine work in very positive ways. We’re not there yet, however. These work platforms shift the risk from the company to the individual who must hustle to maintain a flow of work, find her own insurance, plan for her retirement.
Of course, companies don’t need as many employees as they once did. Consider research from Klaus Schwab, founder and executive chairman of the World Economic Forum, in his book the Fourth Industrial Revolution, that compared the top three companies in Detroit (Ford, GM, Chrysler) in 1990 and the top three companies (Apple, Google, Facebook in Silicon Valley in 2014. The similarities and differences are staggering and, perhaps, foreshadow our future, notably the concentration of capital. The two groups shared similar revenue but the Silicon Valley companies manufacture considerably fewer physical products, while their combined value is four times that of the Detroit automakers. More strikingly, the Silicon Valley companies earn ten timesthe revenue with one-tenth as many employees. (Incidentally, that employee count doesn’t even consider the supply chain of the automobile industry, which is largely non existent in the digital realm where customer (volunteer) contributors freely supply content to Facebok and Google.)
The Next Jobs to Go: Knowledge Jobs (or Why You Should Pay Attention)
This productivity paradox that first affected the hard hat jobs the American Rust Belt is now creeping into the so-called “knowledge worker” jobs that afforded comfortable upper middle-class incomes. More worrisome for many workers, this fourth wave of automation – sometimes called the Fourth Industrial Revolution – is hitting harder and faster than previous disruptions, and will affect cognitive and physical labor just as severely. While computerized cognitive labor is replacing humans, it is also liberating humans to do other, perhaps more creative, work. Much of this new work is yet to be imagined, yet it will no doubt enable humans to reach new levels of potential.
This notion of the Fourth Industrial Revolution is not off on the horizon; it’s happening today as companies use algorithmic and machine intelligence to more efficiently do the work of humans. Consider, for example, the rise of the chat bot taking on sales and support tasks, the potential of autonomous vehicles to replace 5 million Americans currently employed in transportation jobs, and the emergence of the employee-less store. In October 2016, Budweiser transported a truckload of beer 120 miles with an empty drivers seat. In December 2016, Amazon announced plans for the Amazon Go automated grocery store in which a combination of computer vision and deep learning technologies track items and only charges customers when they remove the items from the store. The first Amazon Go opens in Seattle later this year. In February 2017, Bank of America began testing three “employee-less” branch locations that offer full-service banking automatically, with access to a human, when necessary, via video teleconference. JP Morgan Chase & Co. just announced a software product they created and deployed called COIN for Contract Intelligence, which interprets commercial-loan agreements in seconds with fewer errors than humans. This work used to consume, 360,000 hours of work every year by lawyers and loan officers. Rising automation is indeed creeping into the white collar world and eating jobs.
More and more of these automated systems deliver great convenience to consumers and businesses, but at the loss of jobs. In fact, an Oxford Martin in 2013 report predicted that 47% of current work could be replaced by computerized automation like chat bots and driverless cars by 2033. That is just 16 years away! And a concerning recent update to that report suggests the vulnerability to this disruption of work is not evenly distributed geographically and that the potential for job replacement by automation appears to have a direct correlation to education levels. The notion of massive job loss to automation is highly contested, this recent article in the economist summarizes the debate.
These changes are happening right now. Texas, for example, used to offer reliable employment and middle-class paychecks for workers in the oil industry. Then, not long ago, oil prices plummeted and production pulled back, taking good jobs with it. Now, as oil prices rebound, the oilrigs are fired up again, but that job? A machine likely has it. Just ask Eustacio Valaquez, a veteran of the industry who has moved from job to job as technology and automation took on the work he once did. “Pretty soon every rig will have one worker and one robot,” he told a reporter from the New York Times.
In the Rust Belt city of Youngstown, Ohio, pawnshop owner George Beshara remembers when factory jobs allowed his hometown to flourish. “We could use some manufacturing jobs, good paying ones, not these minimum wages ones. When we put tariffs on steel and start manufacturing again, we got a shot,” he told the Guardian’s Chris Arnade.
The right incentives may well bring factories back to the Ohio Valley, and there will be work. Not much of the old style manufacturing work however, will be for humans. That does not mean humans are no longer required, in fact, America is still making things—just differently. We are seeing a surge in demand for highly educated individuals to work in advanced manufacturing. In fact, the U.S. economy will need to fill 3.5 million skilled manufacturing jobs over the next decade, the White House says.
Source: Pulse (Heather McGowan and Chris Shipley)