The United States is the most advanced industrialized nation in the world. It’s economy produces the greatest GDP (total quantity of goods and services), doing so with the greatest productivity (GDP per capita). Rapid productivity increase through technological advance has been, and continues to be, the prosperous economic history of the US. The current wave of technological progress (aka automation, which includes advances in software, semiconductor devices, artificial intelligence, and robotics) is accelerating the pace of productivity increase. That is, while automation is a tremendous gift that allows the US economy to produce more goods and services with less human labor, millions of workers will be displaced. It is incomprehensible, therefore, that the solution to a shrinking demand for workers is, apparently, to increase the supply.
This shrinkage became evident at the turn of the century, when job growth, particularly among lower-paid, lower-skilled, and less-educated workers, fell into stagnation. According to MIT researchers, Erik Brynjolfsson and Andrew McAfee, US job growth has now become decoupled from US productivity growth. They believe that productivity growth will increase relentlessly, as it has since the end of WWII, but US employment growth, which had previously risen with productivity growth until 2000, will remain flat, as it has done so since 2000. The authors attribute the cause of the Great Decoupling to the rapidly declining price of digital labor relative to human labor. The dramatic economic and cultural implications of such automation is examined in their best selling book, The Second Machine Age.
A recent McKinsey Global Institute study estimates that, by as early as 2030, 39 million to 73 million US jobs could be eliminated by automation. There are numerous other studies with similarly ominous predictions (e.g., here, here, and here). The jobs that are most susceptible to automation are ones that involve routine procedures that can be performed by computerized algorithms. These include machinery operation, vehicle operation, fast food preparation, data collection and processing, financial services, and customer services. Well paying jobs that can be obtained today with only a high school education will be the first to go. The jobs that will survive, and thrive, are those involving cognitive tasks, favoring occupations where skill and education are important. They will include engineers, scientists, medical professionals, and teachers. Luckily, they will include jobs for those who apply high cognitive ability and skill to actual work, such as carpenters, electricians, mechanics, plumbers, and welders.
Future jobs, the jobs that will be created by the 2nd Machine Age, will no doubt favor those educated in the STEM (Science, Technology, Engineering, and Mathematics) disciplines. Beyond these job types, it’s anyone’s guess. In a Harvard Business Review interview, Brynjolfsson and McAfee discuss three job categories that will survive automation. The first is “high-end creativity that generates things like great new business ideas, scientific breakthroughs, novels that grip you.” The second is deciphering body language. People who excel at “emotion, interpersonal relations, caring, nurturing, coaching, motivating, leading, and so on” will be in demand. Dexterity and mobility comprise the third category: because “it’s unbelievably hard to get a robot to walk across a crowded restaurant, bus a table, take the dishes back into the kitchen.”
But it’s a pretty good guess that the jobs-of-the-future market will be much smaller than markets that followed any previous waves of technological advance. The number of high-end creativity jobs will be about the same. That is, the number of elites that currently hold such lofty positions is not likely to grow. Even if it did, say doubling, it would still be a statistically insignificant portion of the labor force. Similarly, it’s not likely that we will need a greater number of body language decipherers than we need today. Robot controllers will obviously be needed (e.g., to control fleets of driverless trucks and flocks of pilotless drones). And, of course, the demand for robot repairmen will skyrocket.
Elsewhere, things look bleak. Robots that serve up to 360 (basic to gourmet) sandwiches per hour are looming, and they don’t have to navigate crowded restaurants. Machine operator jobs will no doubt survive. After all, the machines of the future will need human operators. So the guy that ran the old machine that replaced ten workers will be retrained to operate the new machine that replaces 1000 workers. And even in occupations that experience an increase, the gains will be short-lived. For example, AI will eventually replace vaunted robot controllers. Technological advance is simply destroying jobs faster than it is creating them.
Why is it, then, that with a rapidly shrinking demand for labor, we need more workers? In an era of stagnant wage growth, an increase in the labor supply can only cause a decrease in wage rates. The standard answer is that the US economy must be infused with millions of young workers to support America’s aging population in retirement. That is, the working age tax base must be expanded to keep our intergenerational Ponzi schemes (aka Social Security and Medicare) from going broke. America’s declining birth rate, warns Lyman Stone, of the American Enterprise Institute (AEI), “will likely have far-reaching negative economic consequences.” To stave off insolvency, Mr. Stone proposes a “fertility dividend” policy, where the benefits paid to retirees would be based, in part, on the payroll tax contributions of their children. Without some such scheme, “our long-term obligations will have to be financed with substantially fewer people (or, perhaps, substantially more immigrants).”
Our political leaders prefer the latter option: substantially more immigrants, including about 20 million working age adults projected to enter the US by 2030. It’s the only way to increase US GDP growth, say Republican stalwarts such as Paul Ryan, Marco Rubio, Jeb Bush, and Larry Kudlow, and every Democrat worth his open-borders salt. As to the jobs of the future, education reform is their answer. They’ll simply transform the system that has, for many decades, failed many tens of millions of working class Americans, into a system that prepares students and retrains displaced workers for the millions of yet to be determined jobs that they hope will eventually materialize.
There is growing evidence that, over the next decade or two, tens of millions of jobs will be eliminated through automation. Many experts believe that technology has advanced to the point where an advanced industrialized nation can produce economic abundance with a paucity of human labor. Given this possibility, what responsible policymaker can conclude that the US economy needs more fertility and more immigration? Is he willing to bet that tens of millions of new jobs of the future will be created – jobs that will be accessible to working class America? Or that our education system can be reformed to prepare students and retrain workers for jobs whose only currently known characteristic is that they will generally require greater technical knowledge, cognitive ability, and creativity than the jobs that they will eliminate?
Policymakers’ time would be better spent in contemplation of how American society can benefit from the prosperity that, thanks to automation, can be produced by a smaller and smaller labor force. For there is no evidence that a shrinking population is inimical to GDP growth in advanced industrialized nations. A recent study, ‘Secular Stagnation? The Effect of Aging on Economic Growth in the Age of Automation’, published in the American Economic Review, found “no negative association between aging and lower GDP per capita.” As one might expect, advanced industrialized economies with aging populations are the most likely to adopt automation. Indeed, “when capital is sufficiently abundant, a shortage of younger and middle-aged workers can trigger so much more adoption of new automation technologies that the negative effects of labor scarcity could be completely neutralized or even reversed.”
In the US, where capital is abundant, automation relentlessly propels GDP growth. If the projections of the Mckinsey Institute, et al are accurate, then by 2030, the US will produce more GDP with as few as 118 million, possibly as few as 84 million, workers – a truly remarkable feat of efficiency. The problem will be that the US labor force will then consist of 225 million working age adults. Those who clamor for more fertility and more immigration should explain their plan for the 39 to 73 million who will be waiting for the new jobs to open.
It had better be creative. At minimum it should include stipends for interim living expenses. One hopes it will include dexterity and mobility training. Nothing will be more important to tens of millions of former truck drivers, burger flippers, and farm workers than being defter than a robot. And those who are not drawn to body language deciphering, should bone up on their STEM skills.
Finally, no plan would be complete without staunch support for our best and brightest. These include many millions of students and displaced workers who will settle for nothing less than the envied high-end creativity jobs – the high-paying ones where you get to work on scientific breakthroughs and novels that grip.