16 June 2018

Self-driving Vehicles and US Employment

America’s workforce and the self-driving future: Realizing Productivity Gains and Spurring Economic Growth from Securing America's Future Energy (SAFE) comments
In the last several years, the development and adoption of autonomous vehicles (AVs) has emerged as a central policy subject, both in the United States and across the world. The vision of a future where vehicles drive themselves has captured the imagination of the public, promising the potential for significant improvements in roadway safety, economic productivity, accessibility, and reducing fuel consumption and accompanying emissions.
At the same time, some have expressed concern about the long-term impacts of the technology, most intensely with regard to the question of the potentially farreaching impacts of the technology on the U.S. labor force. The individual identities of Americans are often intertwined both with the vehicles they drive and their occupations. The potential significant changes on both fronts in the years and decades to come is, understandably, an unsettling prospect for some. To ensure that policy decisions are made on the basis of solid evidence, SAFE engaged us to answer a series of questions that cut to the core of these issues. The questions were:
1. What precedents should we look to in thinking about the impacts AVs will have on society and the economy?
2. What are some concrete examples that illustrate the nature and magnitude of the economic and social benefits that AVs can offer?
3. What will be the medium- to long-term impacts of vehicle automation on the workforce? Upon what will the scale and timing of those impacts depend? What steps can be taken today to ensure the best outcome for both the public that stands to gain from AVs and the workers whose jobs could be impacted?
These questions were selected because of the importance of improving the social impact of the technology, the potential for impacts on the labor force, and the importance of these considerations to policymakers in weighing AV regulation. A deeper knowledge of the broader economic impacts of AVs will help to encourage constructive choices in a resource-constrained world. 
Over the last six months, we divided these questions amongst this group, with a report dedicated to each question. We performed independent and rigorous research utilizing well-accepted methods of economic analysis that culminated in three reports—referred to in this brief as the Compass Transportation report (focused on the question of precedents), the Montgomery AV benefits report (focused on the benefits of AVs) and the Groshen employment report (focused on the employment impacts)—that each addressed one of the questions posed above.
The authors state
Although they are not yet in widespread commercial use, there is intense public interest in autonomous vehicles (AVs). Much of the focus has been on the broad societal benefits this technology can offer. AVs also have the potential to influence society in a way unseen since the invention of the automobile. In addition to dramatically reducing traffic accidents and roadway fatalities, AVs hold the promise of improved mobility—critical for economic growth and quality of life. AVs can dramatically improve the lives of communities underserved by our current transportation system and those most vulnerable to its inefficiencies, namely Americans with disabilities, seniors, and wounded veterans. 
However, some have raised concerns about the potential for AVs to negatively impact workers and exacerbate wealth inequality. SAFE believes that AV-related labor displacement concerns—many of which have been expressed sensationally—must be addressed seriously rather than merely dismissed out of hand or repeated without verification. In response to these concerns, SAFE commissioned a panel of highly regarded transportation and labor economists to conduct a fact-based and rigorous assessment of the economic costs and benefits of AVs, including labor impacts.
The commissioned research painted a detailed outlook for the future economic and labor market impacts of AVs. They found:
• AVs have many of the characteristics of “catalyzing innovations” whose positive impacts are felt broadly throughout the economy. 
• Significant economic benefits from the widespread adoption of AVs could lead to nearly $800 billion in annual social and economic benefits by 2050, mostly from reducing the toll of vehicle crashes, but also from giving productive time back to commuters, improving energy security by reducing dependence on oil, and providing environmental benefits. 
• A study of traffic patterns and job locations found that some economically depressed regions could see improved access to large job markets for their residents through the deployment of AVs. 
• AVs will create new jobs that will, in time, replace jobs eliminated by automation. Strong workforce development infrastructure can both mitigate employment disruption and speed the evolution of worker skill requirements that will contribute to full employment and economic growth. 
• There is significant time before the impacts of AVs on employment are fully realized. Simulations of the impact of AVs on employment showed a range of impacts that would be felt starting in the early 2030s but would only increase the national unemployment rate by 0.06–0.13 percentage points at peak impact sometime between 2045 and 2050 before a return to full employment. 
• The economic and societal benefits offered by AVs in a single year of widespread deployment will dwarf the cost to workers incurred over the entire multidecadal deployment of AVs when measured in purely economic terms. The benefits of AVs are sufficiently large to enable investment of adequate resources in assisting impacted workers. 
• By pursuing a rapid deployment of AVs, combined with investments in workforce policies that seek to mitigate costs to workers and policies that address other risks or costs that might emerge alongside greater AV adoption, the United States can enjoy the full benefits of AVs as soon as possible while simultaneously preparing the workforce for the jobs of the future. 
Economic and Societal Impact Many of the most compelling benefits of autonomous vehicle technology will be intangible or undetectable from modeling designed to capture incremental gains. Any economic estimates of these benefits should be understood as an attempt to capture just a portion of gains from AVs. This conservative microeconomic analysis estimates economic benefits of up to $800 billion per year with full deployment of AVs. Utilizing the projections for AV deployment that SAFE developed, the value of AV benefits through 2050 will likely be between $3.2 trillion and $6.3 trillion. This is a partial estimate looking at a narrow set of case studies—a full estimate would likely be significantly higher. 
A projection of the annual consumer and societal benefits of AVs is in Figure A. The breakdown of these benefits (upon full adoption) is in Table A. 
Accident Reduction: In 2010, the National Highway Traffic Safety Administration (NHTSA) estimated the economic costs of car crashes to be $242 billion per year. When quality-of-life costs are added into the estimate, the total value of societal harm was approximately $836 billion per year. Extrapolating these values based on more recent crash and driving data puts the annual societal cost of crashes at over $1 trillion today. Using a conservative methodology in which we assume AVs would only address crashes resulting from a gross driver error (e.g. distraction, alcohol, and speeding), the annual benefit would exceed $500 billion. Given that human error contributes to over 94 percent of accidents, benefits could exceed this amount. 
Reduce Oil Consumption: Oil holds a virtual monopoly on vehicle fuels, with petroleum accounting for 92 percent of the fuel used to power the U.S. transportation system. By precipitating a shift away from petroleum as the dominant fuel source, AVs can substantially reduce America’s reliance on oil. An analysis of the energy security and environmental benefits of increased EV uptake as a result of AV deployment supports an estimated $58 billion societal benefit. 
Congestion: Crashes are a major source of road congestion and improved safety from AVs and better throughput (e.g. through reduced bottlenecks) could significantly reduce the current costs of congestion: Close to 7 billion hours are lost in traffic and over 3 billion gallons of fuel similarly are wasted every year. 
Improved Access to Retail and Jobs: SAFE modelling of road speeds around specific retail establishments found that the increased willingness of shoppers to travel—even by just two minutes each way—could increase a mall’s customer base by nearly 50 percent in some instances. Additionally, SAFE modeling identified numerous economically disadvantaged localities for whom better transportation options would lead to greater employment opportunities. For a group of four struggling cities (Gary, IN, Benton Harbor, MI, Elmira, NY, and Wilmington, DE), SAFE modeled how increased traffic speeds from AV adoption and greater willingness to travel could impact the number of jobs within reach. An illustrative example is in Figure B.
The Effect of AVs on the U.S. Labor Force
From the automobile to the internet, history has demonstrated time and again that new technologies lead to sizable economic and social benefits in the long run. However, with significant change always comes the specter of potential loss, particularly in the short term. Like many new technologies before it, the public discourse around AVs has witnessed a significant focus on potential downsides, often with considerable exaggeration. However, the potential losses must be balanced with the benefits from highly significant improvements in safety, reductions in vehicle crash fatalities, gains in productivity, reduced congestion and increased fuel efficiency that will result from AV deployment. Indeed, the benefits are sufficiently large to enable investment of adequate resources in assisting those affected.
A study of historical precedents for the impacts of new technologies found a common pattern: Adoption of new technologies improves productivity and increases quality of life. Widely adopted technologies can transform our way of life and improve economic well-being at a national scale. Often, technological progress leads to improved opportunities for workers in the short term; a recent study found that the rise of e-commerce has, on net, improved jobs for high school graduates.1 However, the impacts of those technologies can also present temporary challenges for the workforce, both for employers needing skilled workers, and for workers whose skills may no longer be as competitive in the labor market.
In the absence of concrete estimates, the media and public have a tendency to concentrate on the worst possible outcome. A recent report claimed that “more than four million jobs will likely be lost with a rapid transition to autonomous vehicles.”2 The methodology used to develop this number was simply to count driving jobs in the United States and assume that they would be rapidly lost as AVs deploy. Such assumptions and conclusions lack context, nuance, or grounding in labor market dynamics and the natural cycle of labor force evolution. Using the scenarios SAFE provided for the adoption of AVs, the Groshen employment report modeled the technology’s impact on the workforce. The study concluded that AVs would not lead to the long-term loss of jobs, although some number of workers could experience unemployment and wage losses. As there are far more professionally employed truck drivers than professionally-employed car drivers, impacts would be tied more closely to the adoption of very high automation in trucks (defined as no driver “in the loop” for most of operation). In contrast, partial automation or teleoperation of trucks is not likely to have significant negative impacts on the workforce.
Figure C and Table A contextualize the job loss within a broadly understood metric—the unemployment rate. Relative to a baseline of full employment, the advent of AVs are projected to increase the unemployment rate to a small degree in the 2030s and to a somewhat larger degree in the late 2040s, with a peak, temporary addition to unemployment rates of 0.06– 0.13 percentage points. Table A contextualizes the size of this employment impact with the shock of the recent Great Recession and a previous mild recession. Policy steps to address the evolution of the labor market must ultimately be placed in the context of the broader impacts of AVs in order to ensure the best outcome. Due to the large-scale societal benefits from the deployment of AVs, policies to address labor force issues must carefully consider their potential impact in delaying the deployment and thus the benefits of AVs. 
Delaying the deployment of AVs would represent a significant and deliberate injury to public welfare. Rather than delaying the benefits, policymakers could ensure that the interests of the people who may lose jobs are well protected through effective mitigation programs. Figure D illustrates the importance of balancing these two priorities. It plots both the conservative projected AV benefits and the range of projected wages that will be lost to individual workers due to AV-related unemployment. The range of projected wage loss reaches as high as $18 billion in 2044 and 2045. However, it is essential to note that this goes hand-in-hand with projected social benefits well in excess of $700 billion for each of those years. In fact, not only are the social and economic benefits of AV deployment significantly more than their costs to workers on an annual basis, but the benefits of AVs each year are far greater than the total cost to workers over the next 35 years combined (illustrated by the middle range of this graph)