While robotics and artificial intelligence (AI) promise great advances in productivity, mostly they seem to worry people. Commentators talk and write endlessly about how these marvelous technologies will steal jobs from both workers and the managerial class, creating a large unemployed population. If history has anything to say, however, and it does, such fears are not only exaggerated, they are off the mark entirely. Ultimately, AI will create more new jobs than it destroys and likely in occupations heretofore nonexistent.
Popular commentary on this matter maintains an almost universally downbeat tone. A January 2018 Gallup poll concluded: “Economists agree” AI is the “single biggest threat to future job growth.” Such thinking, especially within the tech community, goes on to envision the rise of large class of unemployable people. It almost always concludes that the nation must care for these people and protect social cohesion by providing a universal basic income (UBI) to all, financing it with a tax on the great wealth created by AI. Depending on who is speaking, the tax would fall either on robots themselves, their users, or their producers (but seldom the person calling for the tax.)
Before framing policy around such thinking, all should note that the nation has seen this movie before. In the 1930s, a group called the Industrial Workers of the World (IWW) published a report blaming the wide spread unemployment of the time on the use of machinery in production. Its analysis noted how in the prior 20 years the introduction of machinery into factories had cut the labor hours needed to produce an automobile by three quarters and the labor required to produce a ton of steel by more than 85 percent. It further noted how in the prior 75 years, the labor required to produce the wheat crop of the United States had fallen to a mere hundredth of what it had been. The IWW thinkers concluded that those who had benefitted from these surges in productivity should provide incomes to the workers displaced in the process.
If the IWW’s analysis looks ridiculous against the great growth in employment that followed after World War II, new calls of this kind nonetheless arose right in the midst of that great prosperity. In the early 1960s, a large group of academics, including several economics and scientific Nobel laureates, issued a report that identified the “new kinds of automation” as having “broken” the once-secure “link between jobs and incomes.” Concerns along with these lines were in fact widespread. At about the same time as this public-spirited report appeared, President John F. Kennedy warned how automation, among other things, pointed to a future haunted by the “dark menace of industrial dislocation, increasing unemployment, and deepening poverty.” He created an Office of Automation and Manpower in the Labor Department to address the “major domestic challenge of the Sixties,” which he described as the effort to “maintain full employment at a time when automation, of course, is replacing man.” To help, he recommended that Congress fund “readjustment allowances” for “workers displaced by technological change.” Later in that decade, President Lyndon B. Johnson brought together a panel of experts to study this matter. It concluded that the government should create “a guaranteed minimum income for each family.”
Apart from the changes imposed on language by political correctness, such concerns and conclusions sound exactly like those uttered today. And no doubt they seemed just as compelling at the time. Yet, the widespread unemployment forecast at these and earlier technological waves always failed to become a lasting part of the country’s social landscape. To be sure, each period of innovation destroyed jobs. The railroads put out canal builders and workers. The rise of the automobile not only destroyed the work of buggy makers but also what in the early twentieth century were tens of thousands of jobs and millions of acres dedicated to the breeding and training of the millions of horses then needed in the economy, shipping them and selling them. It also ended work for the thousands, possibly millions more stabling and caring for them. The invention of shipping containers in the 1950s put millions of longshoremen out of work. But in each instance, new jobs emerged.
Despite the concerns omnipresent at each innovation, jobs and wealth creation have typified technological waves, going back no doubt to the invention of the wheel and the breeding of animals for draft and transportation. The statistics, sadly, do not go back as far as the wheel, but as far back as economic historians can take them they verify this fact. For all the technological advancement over the centuries, economies, except for brief interludes, have reliably provided work for some 95% of the population that wants to work. If technology destroyed jobs, that number would have fallen over time.
Often the productivity itself creates the additional employment. The application of the spinning jenny and machine loom in eighteenth and early nineteenth century Britain put hand weavers and others out of work. By creating a much more profitable trade in textiles, however, the inventions brought about such an expansion in the industry that it ultimately employed more people than previously and indeed a bigger part of the nation’s workforce as well. More recently, the advancement of automatic teller machines (ATM), by making everyday banking so much more efficient, enabled the industry to expand and employ greater numbers of men and women, including the tellers that people feared would wind up on the unemployment line.
Of greater significance in the employment equation is the wealth created by each technological advance. To be sure, each wave creates a class of the super-wealthy lucky or smart enough to have made themselves part of it. The great railroad barons of the nineteenth century stand as examples, as do today’s computer-based tech barons. No doubt AI will create its own class of this sort. But the technology also drives the general economy, creating new demands for products and services of all kinds that in turn create new jobs. The job-producing leverage is evident in a simple thought experiment. If AI could bring the growth rate of labor productivity in this country from the present pace of about 1.0% a year up to the 2-3% pace averaged in the 1960s, 70s, and 80s, the overall economy would expand at something close to 3.5% a year instead of the recent 2.0% pace. Over the next ten years, that difference would produce $3.6 trillion more annual national product than otherwise, a cumulative addition about equal to the entire federal debt.
Of course, each generation, though it has access to this history, worries that its technology is somehow different. The response is understandable. People can see the jobs the new technology will destroy a lot clearer than the jobs it will create, some of them entirely new. It is noteworthy in this regard that the 1975 report of the Council of Economic Advisors (incidentally authored by Alan Greenspan) did not once use the word “computer.” When, for instance, containers put longshoremen out of work in the late 1950s and early 1960s, it would have taken a bold forecaster indeed to predict that containers would, as they did, foster a 2,000% increase in world trade in just five years and create more jobs for higher-paid, better trained workers in the industry and beyond it. When in the 1980s, the widespread use of word processing drove out the typing pool, mostly women, still the number of women in the workforce increased, as did the proportion of the female population in paid employment.
None can doubt that robotics and AI will cause dislocation and hardship for some. In this case, the nature of the new technology will extend the pain to occupations once thought immune (which may be why fears have become especially intense among journalists and bureaucrats.) There is every reason to seek programs and funds that can retrain and such people for other work. Though, as always, it is difficult to identify what that future work will involve (and a fool’s game to try to identify it in a place like this) the evidence of history, a small part of which this article reviews, and the logic of economic growth suggests that a jobless future is the least likely of outcomes and that the need for a UBI misses the point.