The voices of Urban Institute's researchers and staff
February 19, 2018

Automation’s effects on jobs are more complicated than you might think

February 19, 2018

As artificial intelligence and technology have advanced, Americans have become increasingly preoccupied with how automation will affect workers. If robots can do the same jobs as humans without needing a paycheck or a lunch break, will workers be displaced and pushed into low-quality jobs that are (for now) beyond the reach of automation?

Level-headed analysts point out that automation is a mixed bag: workers could lose jobs in sectors that automate, plenty of new work will be available, and we’ll all benefit if goods and services are produced more efficiently. This analysis is fair, but as Uber Advanced Technologies Group (Uber ATG) recently pointed out, the effects of automation could be complicated and could expand job opportunities for workers in automating sectors of the economy.

Uber ATG is the division of Uber that works on self-driving technologies, including self-driving vehicles in the trucking industry. Its recently published article “The Future of Trucking” predicts how self-driving vehicles could increase trucking jobs by shifting the type of trucking work that humans are responsible for.

Self-driving vehicles are suited to long-haul trucking because highway driving is straightforward and requires fewer split-second decisions and complicated maneuvers. Long-haul truckers also log longer hours than their local trucker counterparts, which raises important health and safety concerns that don’t come up when a robot is at the wheel.

Humans are likely to dominate local trucking jobs, even as self-driving vehicles encroach on long-haul jobs. Uber ATG points out that traffic and efficiency improvements in long-haul trucking because of self-driving vehicles will generate more work for local truckers. Self-driving vehicles would haul loads across the country, and increased activity would boost total trucking jobs as humans deliver locally, creating jobs in an automating sector.

But this could still cause disruption for truckers. Long-haul truckers require a different commercial driver’s license than local truckers, and the license classes required for local truckers are less frequently taught at training centers. Some truckers might need new licenses, and training centers might need to change their programming to accommodate changes. But these new practices and requirements would unfold slowly, and it is unlikely that this retraining would be very disruptive.

Uber is not a disinterested party in these analyses, of course. The company is interested in trucking because of a service it offers called Uber Freight, which helps truckers connect to loads, much like how people connect with an Uber driver to get a ride.

Nevertheless, independent economists have pointed out that even within industries that automate, automation can expand employment opportunities.

In a new paper on artificial intelligence, economist James Bessen explains that consumer responsiveness to price changes helps determine whether automation eliminates or creates jobs in a sector. If demand for a product increases sufficiently as a result of automation, either because the product is less expensive or of a higher quality, that additional demand could employ more workers, even if workers play a smaller role in production.

The Uber ATG and Bessen analyses offer no guarantee to workers. In some industries, automation might decimate the workforce and result in reallocation. But a robust workforce development system, responsive higher education system, and reliable social safety net can cushion that blow. But we don’t need to face the future of work assuming that technological development and workers’ interests are always opposed.

The Georgia Pacific cellulose mill in Brunswick, Ga. Photo by Stephen B. Morton/The Washington Post via Getty Images.


As an organization, the Urban Institute does not take positions on issues. Experts are independent and empowered to share their evidence-based views and recommendations shaped by research.