Abstract
Vehicle automation and other new transportation innovations will reshape the workforce and economy. Examining the impacts of past technological disruptions in other sectors can provide clues about how the likely effects of such disruptions will affect transportation. In this chapter, we consider technological disruptions in farming, manufacturing, shipping, home care, and food preparation. We find that though automating transportation will indeed cause job loss, it will also expand demand for some existing jobs as well as create jobs that are entirely new. The aggregate economic benefits of automation are likely to far outweigh economic declines associated with limited job loss. However, policymakers should take steps now to establish retraining programs and other resources for workers whom automation will displace. Additional research will help pinpoint areas most likely to be affected and better quantify the likely magnitude and timeframe of automation-related impacts.
Overview
Automation is coming to transportation. Exactly how and when is subject to intense debate, but experts agree that sooner or later, it is inevitable. Some believe vehicle automation is a scourge; others believe it is a panacea. Equally uncertain is the impact of this automation on jobs, both for personal mobility and freight.
While automation in transportation is new, automation in other sectors is not. Historical precedents can provide clues about how automation and other technological disruptions in transportation will likely affect the workforce and economy. In this chapter, we explore four instancesâin manufacturing, farming, shipping, and food preparationâin which automated labor-saving devices brought deep structural changes to employment and work. These case studies can and should inform preparations and expectations for the automated-vehicle (AV) revolution.
We find many reasons to be optimistic about the net economic and labor effects of vehicle automation. With passenger travel, we know that automation will displace many driversâfor taxis, limousines, Uber, and Lyftâbut we also know that AVs will enable workers to more productively use the massive amounts of time currently wasted driving and create jobs at all skill levels. Highly trained professionals such as programmers and data scientists will be needed to develop and optimize AV algorithms. Lower-skilled workers will be needed for customer care, cleaning of cars, and more. New services might be offered in the vehicles, such as personal care, business services, and entertainment. With goods movement, automation of long-haul trucking could increase total freight activity and hence increase demand for workers to load, unload, and stock goodsâtasks that are less easily automated.
But the rapid pace at which automation in transportation is occurring warns against complacency. When automation is introduced over generations (as was the case with farming), there was ample time for workers to adjust. Natural attrition of older workers occurs through retirement, and younger workers can be educated and trained to maximize the advantages of automation. If change comes much faster, society as a whole may still benefit but only at the expense of disruptive localized job loss.
Regardless of the exact labor impact, informed public policy is critical for maximizing positive outcomes of AVs, while minimizing costs. Leaders and decision-makers will need to proactively help workers build skills needed in an automated world. Programs should be established soon to support the workers and businesses that automation will inevitably displace in transitioning to new opportunities. And provisions should be put in place to ensure that the benefits of automation in transportation are equitably distributed across geographic regions and socioeconomic classes. The transportation workforce and economy are changing fast. Society must be prepared to adapt.
History of economic and technological transformation
Through the broad reach of history, technology that helps automate tasksâhere broadly defined as reducing the labor input required for a given outputâhas profoundly transformed our societies and economies. But details matter. Time and scale matter the most. Labor impacts will be most disruptive if change is fast and widespread. But net impacts on jobs are likely to be positive.
Division of labor
For millennia, human beings had only two main jobsâhunting and gathering. They also had only two ways to get aroundâtheir left and right legs. It may seem inappropriate to review ancient history, but doing so reminds us of the incredible power we have to change and improve our lives. More or less everything we consider essential todayâshelter, clothing, mobility, sanitation, health care, and moreâis a product of human ingenuity⌠and automation.
As societies became more stationary and food supplies more stable, division of labor allowed individuals to develop specialized skills and pass those skills onto future generations. Specialization fostered innovation, allowing people to create and improve technologies, trade knowledge with other parts of the world, and collaborate on projects too advanced for any one person to carry out alone.
Specialization remains important today. Research shows that all else equal, countries with low specialization are able to do less with capital investment than countries with high specialization [1]. This makes intuitive sense, as a low-specialization workforce is less able to take active, value-generating roles in new technology that arrives alongside investment.
Technology substituting for labor
Classic macroeconomic models of the market depend on capital and labor, and allow technology to essentially substitute for labor. When technology substitutes for labor, by definition some jobs in that specific application will be displaced. It might be natural to think that this would have on net reduced laborâs share of economic production. Over a century of study, however, this has not been the case [2], as overall growth and new employment in other sectors has more than made up for replaced labor.
Autor and Salomons point out that this has given âgrounds for optimism that, despite seemingly limitless possibilities for labor-saving technological progress, automation need not make labor irrelevant as a factor of productionâ [3]. However, they also find in their recent review that âalthough automationâwhether measured by Total Factor Productivity growth or instrumented by foreign patent flows or robot adoptionâhas not been employment displacing, it has reduced laborâs share in value-addedâ [3]. So, while technology has only created economic surplus on the net so far, there are some reasons to question whether this trend will continue in perpetuity.
Many areas where technology substitutes for labor also require energy inputs. Smil has extensively reviewed the history of energy technology and shown that energy and technology together effectively have a multiplying effect on labor, allowing much more output per worker [4].
The net effects of technology introduction historically are so strongly and unambiguously positive that it is hard to imagine a counterfactual world. These changes have happened over decades or centuries, and so seem to have a diffuse effect that is hard to measure during the transition. Introductions of technology for labor can also cause local harm such as pollution and job displacement. In short, substituting technology for labor often results in indirect but widespread benefits for the many at the expense of direct adverse effects of the few. These situations in general can make it very challenging for policymakers to maximize public good [5,6].
Impact of automation in other sectors
Understanding how labor-saving technologies affected other sectors in the past provides insight into how automation is likely to affect transportation in the future. The introduction of technology in farming, mechanization in factories, and standardization of freight with containers each had transformative effects on the workforce and the economy.
These industrial examples are well-documented cases of economic substitution of technology for labor. Automation is also increasingly present in our daily lives. Ready availability of labor-saving devices may have contributed to a shift away from in-home services.
Lastly, dining out has changed some of an unpaid service (cooking in the home) to a paid one (eating in a restaurant or ordering take out). While this is not an effect of automation per se, it may be instructive as an example for some aspects of automation in transportation.
Farming
For centuries, farming was a heavily manual occupation: tilling, sowing, irrigation, and harvesting were all done by hand. Most farms were relatively small...