Low-Skill Jobs Are Booming, But They’re at Greater Risk for Automation

The other day, The Wall Street Journal detailed how Lowe’s will start to roll out robotic shopping assistants at a California store in November. Robots doing customer service? This is another example of how technology and automation are changing the workplace and, over time, the composition of the workforce.

Low-skill jobs such as fast food prep workers and personal care aides are booming, as they have been since the Great Recession. But a new analysis from USA TODAY using EMSI data and other sources indicates that 70% of low-skill positions have a high risk of being automated in 10 to 20 years, compared to 46% of mid-skill jobs and 8% of high-skill jobs.

MaryJo Webster of USA TODAY combined data from EMSI and Oxford professors Carl Benedikt Frey and Michael A. Osborne to arrive at those estimates. Her story was the the third part of USA TODAY’s “Where the Jobs Are” series, an in-depth look at the U.S. economy and local labor markets.

Webster and her colleagues have previously written about middle-skill jobs in part 1 and the college-educated workforce in part 2, all using EMSI’s local labor market data as the foundation. Here are three big takeaways from part 3, on workers vs. automation:

1. It Helps to Have a College Degree … But You’re Not Completely Safe

The more education you have, the less likely it is you’re job will be replaced by robots. Yet all well-educated workers aren’t immune to technological change.

Computers that can write code could reduce the need for computer programmers. Algorithms that generate content could hasten the decline of journalists and other writers. Advanced machines can do some things in the traditional domain of doctors, like replacing an anesthesiologist in low-risk colonoscopy procedures, as USA TODAY mentioned.

This is all part of the creative destruction that Joseph Schumpeter trumpeted 60 years ago.

The good news: technology doesn’t just destroy jobs; it  also creates the need for new types of workers (e.g., app developers after the rise of the smartphone). There’s also the human touch factor. In fields like computer science and journalism, people can provide higher-quality code or articles than machines, at least right now.

Furthermore, those who have specific skills and do them well will usually find a job, even if some of the tasks they used to do are automated. As Webster wrote:

Andrew Crapuchettes, CEO of EMSI, points to recent changes in manufacturing as a harbinger for other industries. Thousands of production jobs were either automated or shipped overseas. But automation has also created new — better paying, although fewer — jobs to operate, maintain and repair the robots and other computerized technology in U.S. factories.

Crapuchettes and others expect the same type of job replacement to occur in other industries, but it could take decades and require significant changes to workers’ skills.

It’s a transition that will be painful, he says. “But when you come out of it, humans tend to figure out a way to survive, and they usually end up in a better spot because of it.”

2. It’s All About Costs

What’s the incentive for investing in labor-saving technology? It comes down to costs. Companies aren’t in the business of adding more workers; they want to be profitable and grow.

Technology can provide better efficiency and quality with menial tasks, as Webster noted. But employers are less likely to invest in that technology if there aren’t a high volume of workers to replace, or if it’s more expensive than really cheap labor.

Food service and cleaning — where the pay is typically at or below $10 per hour — have mostly avoided technological innovation, while higher-paying jobs in factories have been replaced by robots.

“It’s already feasible … for a lot of those low-skill jobs to be automated, at least in part,” says Brian Points, an economist with EMSI. “But the infrastructure and machinery required to make that happen is probably more expensive than to continue paying people that $12 per hour.”

3. Areas Most Likely to Be Hit by Automation: Las Vegas, Corpus Christi, Lancaster, Pa.

Las Vegas has a huge share of casino workers, hotel staff, and others that are at a high risk for automation. This puts Vegas at the top among the 125 largest metros with 54% of jobs in danger of being automated. Right behind it is Corpus Christi, Texas, and Lancaster, Pa. The 10 metros with the largest share of at-risk jobs tend be tourist magnets like Vegas and Cape Coral-Fort Myers, Florida, or metros dominated one or two at-risk industries.

Conversely, cities with high levels of education attainment have the smallest share of high-risk jobs. This list is led by Durham-Chapel Hill (36%), the Washington, D.C. metro (38%), San Jose (39%), and Boston (41%).

Ultimately, though, there’s not a wide chasm between metros on the top 10 of either list, except for a few outliers. The 10th-ranked metro for the highest percentage of high-risk jobs (Greenville-Anderson-Mauldlin, S.C., at 51%) is only slightly better than the 10th-ranked metro with the lowest percentage of high-risk jobs (Worcester, Massachusetts, at 43%).

See more of EMSI in the news on our press page. For more on EMSI’s employment data—available at the county, MSA, and ZIP code level—or to see data for your region, email Josh Wright. Follow EMSI on Twitter (@DesktopEcon) or check us out on LinkedIn and Facebook. And be sure to sign up for the EMSI newsletter.