Automation: Nothing to See Here?
Trickle down robot economics. That’s what we’ve arrived at. Washington Post columnist David Ignatius’ “The Brave New World of Robots and Lost Jobs” set the stage, laying out very well the case for further examining the implications of automation and computerization on the broader economy—a case, coincidentally, that Future Left has been recently advocating for as well. In response, Steven Pearlstein published a rejoinder in the same publication “Workers, Don’t Fear the Robot Revolution.” There’s much I disagree with in Pearlstein’s piece, but allow me to be clear upfront: workers should fear neither robots nor the advancement technology—absolutely not. What is certainly worthy of concern, however, is the philosophical assumption underpinning Pearlstein’s argument: that there is nothing to be done, that political and economic policy ought to remain ambivalent to technology-driven transformation and let the chips, guided only by the mystical wisdom of the free market, fall where they may.
Robot Moguls as Job Creators
Ignatius warned of self-driving cars prompting pink slips for millions of long haul truckers while voice recognition software begins emptying offices of their mid-level employees. Pearlstein replied by pointing out that previous iterations of labor-displacing technology have led to an expansion of jobs in other industries, particularly the service sector. For instance, he invites readers to consider that while there may be fewer manufacturing jobs, the number of yoga instructors has increased. He continues:
“I realize such trickle-down economics has gotten a bad name in some circles, but there is some truth in it. The winners from job-destroying technology hire more gardeners, housekeepers and day-care workers. They take more vacations and eat at more restaurants. They buy more cars and boats and bigger houses. They engage the services of more auto mechanics and personal trainers, psychologists and orthopedic surgeons.”
Obviously, there is indeed “some truth” to this notion, but it seems to be ignoring the broader problem of inequality within the system. For instance, though manufacturing jobs offshored to China allowed for cheaper production of goods and new opportunities built on luxury cropped up to capture the additional wealth brought in the US economy. The lingering problem though is that an increased demand for gardeners, housekeepers, and day-care workers, etc. doesn’t effectively distribute the dividends of the more automated economy to which we are inevitably moving.
Similarly, Andy Kessler, writing for The Wall Street Journal, argues that technology will—as it always has—create more jobs than it consumes. After taking a swipe at notions of a “bigger welfare state and universal basic income and other services to coddle displaced workers,” Kessler explains that technology consistently augments rather than obviates human labor. Again, this seems to be true, but if technology allows one employee to do the work formerly done by two or three or ten employees doesn’t this reasonably allow for less people employed in a given industry?
Consider that Tesla’s Gigafactory, an incredible feat of state-of-the-art engineering and robotics, can produce the Model 3 electric vehicles with little human interaction. CEO Elon Musk referred to the factory as an “alien dreadnaught,” and expects the production process to be fully automated within a few years as the company needs the production rate to exceed “people speed.” While there is little doubt that a such remarkable manufacturing process will require the input of highly-skilled engineers, software developers, and others, the sophistication of the technology that enables the Gigafactory goes beyond mere augmentation.
With 16 million square feet, Ford’s River Rouge Complex in Michigan employed 100,000 people at it’s peak in 1939 each with a 160 square feet of space each. Today it employs 6,000 workers, whose proportion of the complex has risen to 2,667 square feet. The Gigafactory, by contrast, is 5.5 million square feet and employs 272 employees who each have a spacious 20,220 square feet. Of course these remaining employees can be said to be greatly augmented, but for many, many others they are simply replaced.
It’s true that the economy as a whole is far better off with more yoga instructors and gardeners—the US economy is far bigger than it was in 1939 despite the changes within the staffing plan at it’s biggest automotive factories. What is different is that along the way economic growth has become wholly divorced from median household income, which is stagnating. According to the Economic Policy Institute, “Since 1973, the median man working full-time, full-year has seen no sustained growth, dropping from $53,291 in 1973 to $51,902 in 2002 and falling further over the 2002-07 recovery and the recession to $50,383 in 2014.” Andrew Mcafee and Erik Brynjolfsson, authors of The Second Machine Age, refer to this phenomenon as ‘The Great Decoupling.’ When factoring in continuous increases labor productivity—also divorced from median family income—a clearer signal of exploitation is hard to come by:
This graph is evidence that Americans are entrepreneurial, industrious, and indeed are being augmented by advances in technology. (Though automation is only one among other factors contributing to the trends seen above.) But the dividends of this work are not being distributed equitably. Despite justified faith in the fact that automation and artificial intelligence will contribute to continued gains in productivity and growth in GDP, it is clear that there is no guarantee these gains will be experienced by all Americans. In this context, the suggestion that robot-powered trickle-down economics will allow those displaced from their careers to find solace in the service sector as the gardeners, housekeepers, and personal trainers becomes only more unsavory.
It is difficult castigate Pearlstein for putting the onus on individuals to find their own answers to the dilemmas presented by automation. Indeed this is the philosophy of traditional capitalism. “People who lose their jobs must have the willingness and wherewithal to find new opportunities, learn new skills, move to new cities—and to the degree they do not, the economy’s ability to keep everyone employed will be frustrated.” However, the promise of automation is not that middle and lower class Americans will be able to subsist on the lower wage jobs that—for now—can’t be automated. The promise of automation is a paradigm shift in the way we think about work, in the way we approach creativity and therefore innovation more broadly.
It’s a Paradigm Shift—or It isn’t
One way to think automation’s potential to impact jobs is to use, as Abhas Gupta for does using the figures below, a four quadrant chart defined by repetitiveness vs. complexity in terms of labor, those jobs falling in the quadrant of non-repetitive and very complex being those most insulated from the “threat” of automation. While increasingly technology has been encroaching on those three other quadrants, there is no reason we should assume 1) that the fourth quadrant is eternally safe or 2) that everyone has the opportunity—or should have the imperative—to secure one of those coveted fourth quadrant jobs. But ultimately whether advances in technology are a threat or a tremendous benefit to our economic livelihoods is a choice.
If increased automation is a paradigm shift—and I believe that it is—then we should accept such trends as transformational, considering carefully what those transformations can and should mean for society more broadly. Thus it is imperative that policymakers and the public alike bear in mind these trends when discussing the tax code, the social safety net, higher education and continuing education programs, et cetera ad infinitum. Again, this is not the perspective an anti-tech luddite. The “free-market” denial that this technology prompts serious economic and social consideration is not an astute observation, rather it is a choice that society will make to blind itself. It is in everyone’s best interest to ensure we choose more wisely.