(Ed. Note: The “Innovation Fact of the Week” appears as a regular feature in each edition of ITIF’s weekly email newsletter. Sign up today.)
In recent years, it has become almost an article of faith that the world’s economies are facing a tsunami of job destruction from new technologies, including artificial intelligence and robotics.
Martin Ford, author of the jeremiad The Rise of the Robots, warns of “75 percent unemployment by 2100.” Not to be outdone, tech policy gadfly Vivek Wadwa prognosticates that 80 to 90 percent of U.S. jobs will be eliminated in 10 to 15 years. But why settle for 75 percent or even 90 percent when you can pronounce all jobs dead, as Brookings’ Mike Rettig has intoned in a morose piece titled “Will the last human worker please turn out the lights?”
You can trace virtually all this automation alarmism to a 2013 study by Oxford University researchers Osborne and Frey, which warns that technology will destroy 47 percent of U.S. jobs in the next 20 years. Read virtually any article on technology and jobs and you can be all but assured that it will cite this study.
But the Oxford study is just plain wrong. The authors, who didn’t submit their work for peer review, neglected to examine all 702 U.S. occupational categories to manually assess how likely it is that technology will substitute for a human worker in each one. Instead, they took a shortcut: They simply relied on task measures from the Department of Labor, which assessed occupations based on factors such as how much manual dexterity and social perceptiveness an occupation requires. And if their risk score of automation was above 0.7, ipso facto, the job was destined for the trash heap of techno-history.
The only problem is that their methodology produces results that make little sense, as when they predict that technologies such as robots will eliminate the jobs of fashion models, manicurists, carpet installers, barbers, and school bus drivers. Is Versace really going to dress up cute robots in his latest dresses and parade them down the runway? Are we going to be in a Jetsons’ world where you sit down in the magic robot chair and get your hair cut automatically? Is any school district actually going to let a group of elementary school children ride an autonomous school bus with no adult supervision?
It’s worth noting that many articles and pundits also cite a study by the McKinsey Global Institute, claiming that it says that 45 percent of jobs will be automated. But it turns out the McKinsey study actually says that less than 5 percent of jobs could be fully automated. The 45 percent estimate comes from the share of employee time that technology could save—for example, IBM’s Watson cognitive computing system could help doctors make faster and better medical diagnosis and 20 percent of a typical CEO’s time can be replaced by technology like artificial intelligence. McKinsey doesn’t think doctors and CEOs will be out of work because of technology; only that the nature of their work will change and they will be able to use the balance of their time to do more interesting and productive things. Yikes!
ITIF conducted its own manual analysis of these occupations using a very generous assumption about how technology could eliminate jobs, and we estimated that about 8 percent of jobs were at high risk of automation, at most. (For a downloadable Excel spreadsheet detailing the occupations and risk levels, click here.) ITIF derived its “risk of automation” estimates from the U.S. Bureau of Labor Statistics’ (BLS’s) employment projection data series for 2014–2024. We focused on a component of this annually updated data series in which the BLS estimates how many workers were employed in each of the 840 federally defined occupations in 2014 and projects how many will be employed in 2024. The series contains additional data points per occupation, such as average wages and entry-level education requirements. The projections are estimated through a demand-side model that makes a number of economic assumptions. In essence, BLS estimates the goods and services that would likely be consumed by the economy in 2024, then calculates the number of workers that would need to be employed in each occupation to produce that level of goods and services all along the supply chain. While the BLS includes a variable that captures technological change in its projections, it doesn’t make this data available publicly, which makes it difficult for anyone to understand which occupations the BLS considers more “at-risk” of being automated.
To provide easy-to-understand figures on what occupations and shares of employment are most exposed to automation, ITIF qualitatively analyzed each of the 840 occupations and assigned a “risk” level to each on a scale of 1 to 5, with 1 being most at risk of automation and 5 being least at risk. We first sorted occupations according to their education requirements under the assumption that low-skilled occupations were more likely to be substituted by technology. Then we evaluated the job scope of each occupation and the presence of technologies that have the possibility to radically alter them. Our eventual determination of a job’s “at-risk” level was weakly correlated with education, with a correlation coefficient of -0.4.
For example, translators require a bachelor’s degree as a minimum entry requirement, but we assumed that advances in live-translation technologies would have a strong possibility of phasing out the need for human translators over the next 10 years, and therefore assigned it as a “high-risk” occupation. On the other hand, personal care aides do not have a formal education entry requirement, but with the personal care industry built around human-to-human interaction, we classify it as “low-risk.”
From our analysis, 8 percent of workers are employed in “high-risk” occupations, 33 percent are in “moderately high-risk” occupations, 16 percent are in “moderate-risk” occupations, 28 percent are in “moderately low-risk” occupations, and 15 percent are in “low-risk” occupations. Combining BLS employment estimates with ITIF’s “risk of automation” variable, we find that employment in “high-risk” occupations will increase 2 percent by 2024 while employment in “low-risk” occupations will increase 10 percent by 2024. Meanwhile, the labor force is expected to grow by 5 percent.
So, we should all take a deep breath. Every time you read another piece about the coming “jobapocalypse,” think about how hard it is to automate jobs like mining engineers, fish and game wardens, dentists, makeup artists, transit police, administrative law judges, museum technicians, dancers, elevator installers and repairers, special education teachers, funeral attendants, therapists, urban planners, advertising managers, legislators, lodging managers—and, yes, even barbers, fashion models, manicurists, carpet installers, and school bus drivers.