America’s next great employment crisis will likely be more serious than the COVID-19 shutdown or the Great Recession. It could possibly be worse and longer lasting than even the Great Depression. It will not be triggered by disease or a business bubble bursting. Ironically it will be caused by an unprecedented increase in productivity triggered by the increased use of robotics and artificial intelligence (I will use the acronym R/AI here for brevity). Unlike 2008 and 2020, this will not be a crisis caused by any single event. It will be a rolling series of job displacements that will eventually impact millions. The gradual nature of the job losses will likely pass largely unnoticed at first. The political party which shows it can manage this long-term crisis best will likely be dominant for the next several decades.
The Great Displacement has already begun, albeit quietly. When in Chantilly, Virginia, recently, I visited a convenience store that is totally automated. I went through the entire shopping process without interacting with a human employee. There was one human worker in the store. After finishing my purchase, I asked her what her function was, and she said that it was to help customers if they had problems with the checkout machine. Any reader who has visited Walmart recently has probably noticed that there are very few checkout lanes manned by humans. This is not accidental. Automated checkout is the wave of the future. Walmart is also reportedly looking to automate its shelf stocking functions as well.
Conservatives have a bad habit of letting the progressives come up with a solution and then trying to shoot it down. This is an opportunity to get ahead of the power curve and force the left to react.
Robots and Artificial Intelligence (AI) are not automatically interchangeable terms. A robot is a machine that can carry out tasks automatically. Robots have been around for a long time and have been on production lines for decades. Artificial Intelligence (AI) is a computer program that can replicate human decision-making. Combining the two into R/AI is relatively new, but it has made great strides in the last few years. R/AI vehicles have been the first manifestation, and the first large scale displacement of human drivers will likely be in the trucking business, particularly long-haul trucking.
R/AI does not tire, nor does it require federally mandated rest periods. It does not join unions or demand higher pay. Once R/AI is certified for autonomous operations on the open road, R/AI will be irresistible to trucking companies worldwide. To give an example of the scope of the challenge, there are approximately 3.5 million licensed commercial drivers in the country and most of them will face displacement in the next decade.
Uber, Lyft, and taxi drivers will also be on the chopping block. R/AI does not sexually harass passengers or drive distracted. Fleets of automated public transport are likely the wave of the future. Busses and subways will likely still be required to have a standby human aboard to avoid mass casualties if R/AI fails. It will not be long before taxi companies buy fleets of autonomous vehicles and run Uber and Lyft out of business. The numerical displacement estimate for humans runs in the range of 1.5 million. The impact will likely not be as severe as in trucking because most Uber and Lyft drivers have a primary career to fall back on.
Air transport will not be immune. We can expect FedEx, UPS, and Amazon as well as other companies that maintain their own aircraft fleets to fully automate by the end of the decade. As with public ground transportation, commercial airlines will probably be required to have a qualified human pilot aboard for safety reasons, but we can expect the number of the 41,600 commercial airline pilots now on duty to be reduced by at least half.
The fast food and convenience store industries will likely turn to AI near simultaneously with their contemporaries in transportation. The fast-food industry averages 3.65 million employees today. There is a very real possibility that 90 percent of them could be displaced by R/AI. The same holds true with convenience and grocery store employees: they number roughly 2.68 million. Ironically, the reluctance of post-COVID-19 employees to return to work will likely motivate both industries to speed up automation. As Biden-era cash hand-outs dry up, many workers will find that they have cut their own throats.
White collar workers will not be immune. Perhaps the first wave of displacements will come among paralegals. There are an estimated 333,000 of them. Mobile R/AI units can do the routine paperwork, pick-up and deliver court documents, and even do wills and powers of attorney.
In this column, I have only listed the six most at-risk jobs, and there are many more that could fall in the at-risk category. I have not counted other careers that are potentially at risk but experiments with R/AI in the fields of heavy machinery and nursing home care are being conducted. The six discussed here alone represent a staggering 11.3 million potentially lost jobs. At the height of the Great Depression in 1933, 12.8 million people were out of work. The population is larger now, but the impact of that level of unemployment would still be staggering.
Unlike past economic crises, this one will not be triggered by one catastrophic event; it will be rolling and will not be immediately noticeable to the public. There is good news and bad news in that. The good news is that we can see it coming and will have time to prepare. The bad news is that it will be more difficult to convince people they have a problem until it is too late. During the Great Depression, President Herbert Hoover held out hope until his ouster that things would get better. In this case, there will be no such hope. This will not be a typical depression. Unemployment will be accompanied by unparalleled prosperity and productivity. The challenge of caring for and feeding displaced workers will be unprecedented.
Potential Solutions: There are three potential solutions to the problem. Each has political, social, and economic ramifications. As noted earlier, the political party that chooses one and pushes a solution through Congress will likely become dominant through the next few decades.
The Luddite Approach: The simplest approach would be to ban selected industries from replacing humans with R/AI. This would obviously be favored by big labor, particularly the Teamsters, as long-haul trucking is the industry most likely to be impacted first. Like all seemingly simple solutions, this one has unintended consequences. First, we will become far less competitive as a nation. Russia and China will have fewer compunctions about displacing human workers. Second, such legislation would likely be held unconstitutional by the courts.
The Socialist Solution: The most obvious solution that will come from the progressive Left and Social Democrats will be to tax the pants off corporations that displace humans with R/AI and create a fund for permanently reimbursing displaced workers. This has the primary advantage for the Left in that it will create a long-term underclass of people beholden to the government — read the Democratic Party — creating yet another entitled voting bloc. This would move the Left closer to its cherished goal of a socialist America.
The downsides for the country are numerous. Like the Luddite approach, it will make the nation less competitive and productive compared to our economic rivals; worse still, it will drive many industries that can leave to places more accommodating to R/AI. This will have the economic impact of weakening the tax base which socialism needs to thrive.
Taxing the Robots: The approach that would likely be favored by Republicans and other conservatives would be to tax the robots, or more correctly, the job that the displaced human held. As with the socialist approach, it would be placed in a fund that would keep the fired workers afloat until another job could be found, but it would also fund job retraining for a skill not at risk of R/AI displacement. The nation needs millions of skilled carpenters, plumbers, mechanics, welders, and other building trades. The company would be taxed at the rate the displaced human was paying before getting canned. The difference between taxing the robot and taxing the corporate profits is not semantic. The wage paid the displaced worker is a fixed cost embedded in the human’s salary. The company would be gaining that portion of the human salary saved and not paying medical insurance. Thus, it is still ahead of the game.
How would this work in practice? Let’s take the situation of long-haul trucker John Smith. He was making $60,000 and paying taxes in the 22 percent bracket a year before an R/AI vehicle displaced him. The fund reimburses him $40,000 a year while he is unemployed, but he pays no taxes while on the dole. There is also a carrot to get him reemployed. If Smith enrolls in a job training program, he will get the full $60,000 while in school, but he will pay taxes at 22 percent. If he enters his new profession at the entry rate, he will continue to get assistance from the fund until his new salary reaches the $60,000 level; at that time, he leaves the program.
Everyone wins here. The company and nation will profit through the increased efficiency and productivity of R/AI. Smith will continue to be able to feed his family and pay his rent or mortgage while being incentivized to train for a new career.
No plan is perfect. This one will have several downsides, but they can be self-correcting. First, it will not pay for itself immediately. However, as the Smith’s of the nation become reemployed, they will be paying taxes. In the long run, the tax base will increase because Smith and the robot are both paying taxes. This should make conservatives happy because it will be the only government program that expands the tax base without raising taxes; this will please the IRS. Second, human nature will dictate that there will be some happy to stay on the dole and accept the $20,000 pay cut. This can be compensated with a stick. Let’s say Smith’s neighbor Jim Jones remains on the dole. If he does not enroll in a retraining program or find a new job in six months, he will begin paying taxes on the $40,000 stipend. It will then get less comfortable to be a lay-about.
Even hardcore conservatives like Herbert Hoover came to realize that in severe situations, some government intervention is needed to prevent humanitarian catastrophe. But as Franklin Roosevelt found out in the New Deal, Keynesian economics can only apply a bandage on a sucking chest wound. It took World War II to get us out of the Great Depression. This approach holds out the possibility of a long-term solution to a long-term problem.
Conservatives have a bad habit of letting the progressives come up with a solution and then trying to shoot it down. This is an opportunity to get ahead of the power curve and force the Left to react. Republicans should have such a Robotic Relief Program (RRP) in their quiver, but now is not the time to roll it out. In politics, timing is everything. The Republicans should wait until the first round of lay-offs — probably in long haul trucking — before laying out their program. At that point, the public’s attention will be riveted on the problem, at least for a short time. Then, the Republicans should introduce RRP with the message: “You could be next.”
Gary Anderson was the Chief of Staff of the Marine Corps Warfighting Laboratory which did early pioneering work with military robotics and AI.
Notice to Readers: The American Spectator and Spectator World are marks used by independent publishing companies that are not affiliated in any way. If you are looking for The Spectator World please click on the following link: https://thespectator.com/world.