A Quarter of All U.S. Jobs Are at High Risk of Being Automated, Report Finds
A full quarter of all U.S. employment is highly vulnerable to automation, a new report from the Brookings Institute finds. That’s the equivalent of 36 million jobs. Meanwhile, some 36 percent of U.S. employment—52 million jobs—will “experience medium exposure to automation in coming decades.”
While the remainder of U.S. jobs are likely to experience only low levels of automation, the report finds that “Almost no occupation will be unaffected by the adoption of currently available technologies.”
That’s not quite the tagline I’d give the chart below—I think I’d go with “Most jobs are reasonably-to-very susceptible to automation,” but hey.
These findings aren’t particularly shocking, but the report does a nice job of breaking down some of the granularities of where, how, and who automation will hit hardest.
For one thing, rural areas will see more automation risk than urban ones—largely because manufacturing and agriculture work is concentrated there, and rural economies are less diversified than metro ones. (Note that really rural areas and very small towns are less vulnerable—it’s harder to automate many local businesses and tiny, non-agriculture operations.)
Low-paying jobs are the most vulnerable, as you’d expect, because they tend to be the most routine, and thus most automate-able, kinds of work.
Again, rural and Rust Belt states are, as a rule, more exposed to automation risk, for much the same reasons described above. As a side note, there’s a ~pretty~ strong corollary between Red States and automate-able states—many have been subject to a legacy of mechanization and outsourcing—which continues to support the theory that automation helps fuel resentment and class anger and, well, Trump.
Hispanic, Black, and Native American populations are significantly more likely to see their jobs automated than white or Asian populations—again, due to structural inequalities, minorities are more likely to be stuck with lower-wage, sh***y, routine jobs.
I do not agree with the policy recommendations of the report, which lead off with the vaguely dystopian “promote a constant learning mindset”—as in, expect to continually try to adapt to the ever-accelerating rate of automation or else—and continues on to some very mild suggestions for reforming income supports for displaced workers. It suggests a “Universal Adjustment Benefit” which seems to be a nod towards a Universal Basic Income, just much sh*****r and more conditional. It’s basically glorified career counseling: Brooking recommends policymakers automatically enroll displaced workers in counseling, combined with job training programs, and, ideally, more robust income support (which, admittedly, would be good).
I’ve said it before and I’ll say it again, but I’m very tired of these reports casting automation as a force of nature, and then catering to the needs of the companies, owners, and organizations doing the automating. The policy suggestions almost all entail trying to ensure workers can keep up with the accelerating pace of production, that they be able to maximize profits and productivity for companies whatever the scenario.
Where is the report that suggests workers themselves benefit even one iota from all this automation—say, in the form of increased benefits now afforded by the greater profit margins generated by automation? Or in the form of, gasp, a shorter working week? There’s a policy recommendation that makes sense in the age of mass-mechanizing work—and that’s always been the dream of automation, right? Why not look across the country and see all these tasks getting automated and propose a strategy of coexisting with the robots that are taking our jobs? The short answer is that this does not benefit the executive and managerial class. And if workers are going to see the benefits from automation, as usual, they’re going to have to fight for them.
Interestingly, a movement is beginning to form on precisely that front, and a bit later this week, I’ll have news of one of its initial volleys.
