The Covid-19 pandemic accelerated existing trends toward workforce automation as employers sought to minimize labor costs and new work-from-home imperatives increased digitalization of workplaces across many sectors. Putting pressure on employers to automate jobs were factors including competition for workers amid high employment rates, spread of illness among workers in close physical proximity to each other, and political and market pressures to bring strategically significant manufacturing jobs back within U.S. borders. Employees themselves helped drive the push toward remote work, with almost half of workers being able to do so during the pandemic’s early months—and significant proportions of them continuing to do so now, more than three years in.
While there has been considerable research into the ways that automation may eliminate or diminish the number of nonspecialized jobs, much less attention has been paid to how this vocational displacement may interact with continuing structural issues of racial inequality in the United States. In our paper, building on previous research done at The Brookings Institution, we examine the pandemic-spurred growth in automation and show that Black and Hispanic workers are overrepresented in jobs most likely to be affected by automation.
Underpinning the discussion is the reality that these workers are also overrepresented in jobs where physical presence at a job site and nearness to other people are necessary and underrepresented in jobs best suited to benefit from the perceived advantages of new work-from-home options.
The Covid pandemic caused rapid, immediate job losses in the United States: about 20.5 million jobs from March to April 2020, by far the largest one-month decline the Bureau of Labor Statistics has ever recorded. Since then, the economy has built back up to its present rate of unemployment, low by historical standards.
But that’s not to say the country is back to a pre-pandemic “normal” or that everything is rosy for the average worker. In between then and now, intersecting trends around consumer behavior, labor markets, and the degree of automation have wrought meaningful change, much of which has fallen on the shoulders of workers earning low and moderate incomes.
Massive increases in online purchasing pushed e-commerce companies, such as Amazon, further in their adoption of and experimentation with automated technologies, including warehouse robots and autonomous delivery vehicles, that seem destined to slowly displace human workers .
The rapidly increasing digitization of work threatens to leave behind those without digital skills or resources. Statistics from the pandemic suggest that in some ways it already has. In addition to the proliferation of online meetings and their attendant platforms, businesses ramped up their use of cloud computing, online contracting, and digital payments systems, among other examples. Working from home was not an equal opportunity: Two-thirds of those with a bachelor’s degree were able to telework, versus one-fourth of those with a high-school diploma, according to a U.S. Bureau of Labor Statistics report. Among White workers, 49% could work from home, compared to 39% of Black workers and 29% of Hispanic workers in the labor force. Ability to telework was much greater for those in higher- versus lower-paying jobs, and, as expected, remote work was much more feasible in the traditional “white-collar” occupations.
Employers , scrambling to find workers amid the many constraints imposed by the pandemic, accelerated their use of automation. Orders for robots in North America, mostly in the U.S., increased by 20% in the first quarter of 2021. At airports and on planes, sanitation robots were deployed to cleanse surfaces and the air using UV light sterilization and air sanitizers. In airports, in senior living facilities, and on college campuses, robots were put to work in food and retail delivery. At grocery stores, robots began processing transactions, cleaning floors, stocking shelves, and delivering groceries.
This is a dynamic situation that will require years of study and the perspective of time to fully understand. But we can begin to take stock of which jobs are endangered by automation and what the demographics are of those who currently hold them.
In our study, we found that Black workers are overrepresented in 17 of the 30 U.S. occupations with the highest risk of being automated (as classified by a 2017 study by Frey and Osborne), and Hispanic workers are overrepresented in 22 of these occupations. Retail salespeople, cashiers, construction workers, laborers, and secretaries are the five categories of endangered jobs that employ the highest number of workers. At the other end of the spectrum, Black workers are overrepresented in just 5 of the 30 positions at low risk of being automated, jobs including early childhood educators and dietitians/nutritionists.
As automation takes hold, it is expected to significantly impact Black and Hispanic workers. For example, in a 2017 study, I found that Black workers were more than one-and-a-half times more likely than White workers to be cashiers, cooks, food servers, production workers, and laborers and freight/stock/material movers. I found that Black workers were more than three times more likely than White workers to be security guards, bus drivers, and taxi drivers/chauffeurs.
Experts have spoken of a coming Fourth Industrial Revolution, one that will be driven by increased automation (including the burgeoning field of artificial intelligence). For all workers to thrive and to avoid increasing racial disparities, policymakers should consider new approaches in education and workforce training that will help prepare this population for the coming changes.
High schools and community colleges may need to beef up their offerings in courses that teach robotics, programming, and other relevant skills, especially in communities most vulnerable to education and job loss. Policymakers could also actively involve businesses, which will need workers to manage automation, in developing such programs and ensuring they are available particularly in vulnerable communities.
Further, because risk of displacement drops significantly for both Black and White workers who hold a bachelor’s degree , investments in higher education, particularly in historically Black colleges and universities (HBCUs) and institutions that focus on increasing access for students of color, can increase educational attainment and lower displacement risk from automation. Such investment might include funding for technical infrastructure, along with initiatives to increase educator-employer connections, such as the engineering program that links Toyota, the University of Kentucky, and the HBCU Kentucky State University.
Education can and should continue on the job. Studies show companies often recapture the costs of training programs through productivity gains from workers whose skills have been boosted. Many companies, however, are reluctant to make this investment. Policymakers could help by offering government incentives such as expanded tax credits for on-the-job education and training expenses.
Another promising area is in working to narrow a racial gap in apprenticeship programs, which combine work-based training and classroom instruction to develop a specific vocational skillset. In 2019, Black workers made up about 10% of the more than 280,000 people completing a Registered Apprenticeship Program, while Black Americans were about 13% of the U.S. labor force overall.
The data show that jobs held by people of color are disproportionately vulnerable to automation. Combatting this reality will require policy responses that pay extra attention to the issue.