Are the Robots Coming for Opportunity Occupations?
The jobs are here today. Will they be gone tomorrow?
Mels de Zeeuw aims to find out. For several years, de Zeeuw, a senior analyst in the Atlanta Fed’s Community and Economic Development team, has compiled data on opportunity occupations, which are jobs that pay at least the national median wage but do not require a four-year college degree.
That work has proven useful, to be sure. Yet de Zeeuw and colleagues have received considerable feedback focused on a relevant question: What is the future of these jobs in the face of rapid automation? In particular, de Zeeuw noted, questions centered on truck driving, among the most numerous opportunity occupations but one directly threatened by the rise of driverless vehicles.
It seemed a natural progression, then, to address the issue of how automation might affect opportunity occupations more broadly.
De Zeeuw did so by meshing the CED’s opportunity occupations data with research examining which jobs across the entire labor market are most and least vulnerable to automation, producing a report on his findings.
For those who hold or pursue opportunity occupations, de Zeeuw’s overall findings are not especially encouraging. In sum, this broad batch of jobs faces roughly twice the risk of elimination by technology as jobs that require at least a bachelor’s degree. Roughly half of all positions in opportunity occupations could be rendered obsolete over the next two decades, de Zeeuw found. And, indeed, truck drivers face about an 80 percent chance of being replaced by an autonomous vehicle, according to the two studies de Zeeuw consulted in his work.
Are they coming for your job? Maybe, but maybe not
Yet the outlook is not uniformly bleak for opportunity occupations. Many health care sector jobs, in fact, rank among those least likely to be replaced by technology, de Zeeuw said. Employment as a registered nurse and licensed practical nurse stand just a 1 percent and 6 percent chance, respectively, of being automated.
Moreover, before surrendering to a hypothetical advancing robot horde, a few large caveats should be considered, de Zeeuw stressed. The main one is that predicting the future is extraordinarily difficult, and forecasts vary greatly of automation’s impact on human employment. Other credible research, he pointed out, paints a far less threatening picture than the more dire forecasts.
What’s more, experience has already shown that where technology eliminates jobs, it can also add them. Consider spreadsheets. The number of bookkeeping jobs has declined because inexpensive software can quickly perform the same work, and the number will likely keep falling, according to the U.S. Bureau of Labor Statistics. At the same time, the oceans of data produced by technology (such as spreadsheets) have created millions of jobs for analysts and data scientists, de Zeeuw noted.
Much remains unknown
One of the reasons the Atlanta Fed studies the future of work is because so much about it remains unknown, de Zeeuw said. Research and quality data are lacking on how we work with machines and the ways “cognitive technologies”—those that mimic human thought, such as artificial intelligence—affect how companies operate, said Stuart Andreason, director of the Atlanta Fed’s Center for Workforce and Economic Opportunity (CWEO). Research and data are especially scant on lower- and middle-skill jobs that typically do not require a four-year degree, according to de Zeeuw.
“It’s our bank’s mission to promote economic mobility,” he explained. “There’s a lot of focus among policymakers on student loans and accessibility to college, and that’s important. But you’re describing a subset of the population whose opportunities tend to be better than for those who don’t have a college degree. I think it’s also very important to focus on the opportunities for the almost two-thirds of folks over 25 in the U.S. who don’t have a bachelor’s degree.”
Only about half of all jobs in the country appear to require a bachelor’s degree. That share is even lower in every southeastern state except Georgia, according to online help-wanted advertising data from Burning Glass Technologies that de Zeeuw analyzed (see the chart).
Given the importance of opportunity occupations, CWEO aims to continue expanding and deepening its expertise in that area. In the short run, the center plans to add the vulnerability-to-automation information to the online Opportunity Occupations Monitor. In the longer run, CWEO will likely pursue a deeper understanding of the training and credentials required for opportunity occupations that are less threatened by automation, de Zeeuw and Andreason said. This line of research also dovetails with the Atlanta Fed’s work on benefits cliffs and career advancement.
“You can’t predict the future,” de Zeeuw said. “But to the extent that we have some indication of how middle-skills jobs will be affected by technological change, we should make sure workers, employers, and policymakers are equipped with the information needed to align skills with the future needs of the labor market.”