There are many sides to the question of how, when, and whether robots will take human jobs. They range from a belief that automation creates more and better jobs to a fear that human beings will be degraded to full-time video game playing in 30 years.
Now research suggests that there is another aspect of this question that we need to consider. Will automation reinforce societal inequalities?
A paper in the Hamilton Project analyzed the racial makeup of workers in the 30 jobs most threatened by automation and the 30 jobs least likely to be taken over by robots. They discovered that the jobs threatened by occupation were more likely to be filled by Black and Hispanic workers, while the less-threatened jobs were most likely to be filled by Asian-American and White workers.
Cashiers, sales clerks, clerical workers, hospitality workers, and laborers are at the top of the most threatened list.
Low-risk jobs include teachers, healthcare professionals, CEOs and upper-level managers, engineers, and clergy.
A few examples show that Black and Hispanic workers are overrepresented in the most threatened jobs:
- 29.5% of cabbies and chauffeurs are Black.
- 25.8% of truckers are Black.
- 28.1% of food preparation workers are Hispanic.
- 32% of servers in restaurants are Hispanic.
13% of the U.S. workforce is Black and 18% is Hispanic.
In contrast, only 5 of the least-threatened jobs show higher proportions of Black workers. For example, 15% of preschool teachers are Black. That is more than the 13% we would expect to see if the distribution were equal in the population as a whole.
None of the 30 least-threatened jobs showed over-representation of Hispanic workers.
Automation, in the absence of intervention, seems likely to reinforce nd perhaps increase inequities for Black and Hispanic workers.
What kind of intervention might help? Education and training are the obvious choices, but these efforts may not work out as we would hope. Will on-the-job training move cabbies into CEO positions? Will clerical workers easily transition into teaching?
This is another of the concerns that we should probably be thinking about now, rather than waiting until the problem is upon us.