“Broadly speaking, many — perhaps most — workplace technologies are designed to save labor,” wrote MIT economist David Autor in his 2015 paper, “Why Are There Still So Many Jobs? The History and Future of Workplace Automation.” “Given that these technologies demonstrably succeed in their labor saving objective and, moreover, that we invent many more labor-saving technologies all the time, should we not be somewhat surprised that technological change hasn’t already wiped out employment for the vast majority of workers?, asked professor Autor.
The answer, he noted, underlines a fundamental economic reality that’s frequently overlooked: “tasks that cannot be substituted by automation are generally complemented by it.” Automation does indeed substitute for labor. However, automation also complements labor, raising economic outputs in ways that often lead to higher demand for workers. Articles often overstate the extent of machine substitution for human labor “and ignore the strong complementarities between automation and labor that increase productivity, raise earnings, and augment demand for labor.”
Most jobs involve a number of tasks or processes. Some of these tasks are more amenable to automation, while others require judgement, social skills and other human capabilities. But just because some of the tasks have been automated, does not imply that the whole job has disappeared. To the contrary, automating the more routine parts of a job will often increase the productivity and quality of workers, by complementing their skills with machines and computers, as well as enabling them to focus on those aspect of the job that most need their attention.
The advent of automated teller machines (ATMs) in the 1970s is a case in point. By 2010, there were approximately 400,000 ATMs in the US. But, not only were bank tellers not eliminated, but their numbers actually rose modestly from 500,000 in 1980 to 550,000 in 2010, driven by two major forces. First, as a result of IT advances that reduced the costs of operating a bank branch, the number of urban bank branches increased significantly by 2010. Second, while most of the routine cash-handling tasks of bank employees were automated by ATM technologies, IT-based innovations enabled a broader range of bank employees to become involved in relationship banking, providing a variety of relationship-based services to customers, including credit cards, personal loans, mortgages, and investment options.
“A vast economic literature analyzes how rapidly evolving digital technologies — information and communications technologies, robotics, artificial intelligence — affect employment, skill demands, and earnings levels,” wrote professor Autor along with economists Caroline Chin, Anna Salomons, and Brian Seegmiller in a recently published paper, “New Frontiers: the Origins and Content of New Work, 1940-2018. Most of this empirical research has focused on the decline of existing employment in advanced economies due to the impact of automation on mid-skill blue- and white-collar occupations, they added. However, this research has been relatively silent on the countervailing emergence of new occupations and job categories in response to technological innovations that complement and augment human labor.
To shed light on the origins of new occupations and job categories, the authors analyzed the emergence of new work in the US between 1940 and 2018, that is, the introduction of new job tasks requiring specialized human expertise. To conduct the study, they constructed a database of new job titles over those eight decades based on the US Census Bureau job descriptions of new occupations and industry categories in each decade.
This is a rather long and complex paper to summarize because it covers so much empirical and theoretical ground in its nearly 120 pages. Let me briefly discuss a few of the paper’s key findings.
The majority of current employment is in new job categories introduced after 1940. New work has shifted from middle-paid production and clerical jobs in 1940-1980 to high-paid professional jobs and low-paid service jobs since 1980.
As of 2018, 60% of employment are in new occupations introduced since 1940. In the first four decades, 1940-1980, a major share of new work was in mid-skill, mid-wage occupations, such as blue-collar production and operation jobs and white-collar administrative and clerical jobs. Many of these jobs dealt with the kinds of routine physical and information tasks that can be well described by a set of rules and have thus been prime candidates for technology substitution as well as for offshoring to lower-cost countries.
In the more recent four decades, 1980-2018, new work creation has primarily shifted to high skill, high wage jobs requiring the kinds of expert problem solving and complex communications skills typically seen in managerial, professional and technical occupations, and to low paid jobs that generally involve physical tasks such as protective services, food and cleaning services, personal care and health care aides.
This pattern of new work creation and automation helps to explain a major portion of the job polarization seen in the US since 1980.
Technological innovations both automate existing work and instantiate demands for new expertise. Some technologies do both at the same time, but these features are distinct and measurable.
New work emerges in response to technological innovations that lead to new processes, products, services, and to entirely new industries that create demands for new occupations and expert knowledge. The paper showed how new technologies and patents lead to the creation of both augmentation and automation innovations by using natural language processing tools to map the full text of all US utility patents issued between 1920 and 2018 to the titles and descriptions of new occupations created between 1940 and 2018. Each patent was classified as an automation innovation, an augmentation innovation, both, or neither
“In our terminology, augmentation innovations are technologies that increase the capabilities, quality, variety, or utility of the outputs of occupations, potentially generating new demands for worker expertise and specialization. Conversely, automation innovations are technologies that substitute for the labor inputs of occupations, potentially replacing workers performing these tasks."
Augmentation innovations spur the emergence of new work since they both create new tasks that raise the demand for labor and increase the value of occupational services. However, automation of labor tasks reduces employment by displacing workers from existing tasks rather than directly creating new work tasks. In addition, occupations that are more exposed to augmentation are on average also more exposed to automation, a finding that’s not surprising given that many technologies contain both automation and non-automation components.
New work is not merely a technological phenomenon. Much new work is also driven by changing incomes, tastes, and demographics, such as services for an aging population.
Another major source of new work is so called positive demand shocks that drive prices higher and create incentives for entrepreneurs to introduce new automation and augmentation innovations. Positive demand shocks spur the introduction of investments that cause sectors to expand, and thus create not juts more work but also more new work. Similarly, negative demand shocks don’t merely lead to less work, but slow down the introduction of new work in the affected sectors, as has been the case with the impact of the China trade shock on manufacturing jobs.
The paper highlights two salient attributes of new work. “First, new work typically requires expertise acquired through study, apprenticeship, or practical experience (e.g., in cosmetology, in the electrical trades), though the extent of expertise clearly varies across categories. Second, new work often reflects the development of novel expertise within existing work activities (e.g., electrical trades skills specific to solar installations) or through an increase in the market scale of a niche activity (e.g., nail care) rather than a fundamentally new human endeavor.”
Finally, most technological innovations are not primarily about automation but about expanding the domain of human capabilities. Airplanes didn’t automate existing forms of human flight, nor did electricity automate existing forms of light or energy.
In conclusion, the authors suggest some important questions that future research may seek to address, including:
Is automation accelerating relative to augmentation, as many researchers and policymakers fear, or is it primarily the case that the relative balance of automation and augmentation has shifted across occupations and skill groups?
“On the latter possibility, our evidence is clear: automation has intensified in middle-skill occupations while augmentation has concentrated in professional, technical, and managerial occupations and, to a lesser extent in personal service occupations. On the former possibility, the results above suggest that in recent decades, the demand-eroding effects of automation innovations have intensified whereas the demand- increasing effects of augmentation innovations have slowed.”
Is ‘new work’ more labor-augmenting than ‘more work’?
“The finding that augmentation innovations increase occupational wagebills by boosting both employment and wages suggests that ‘new work’ may be more valuable than ‘more work’ — plausibly because new work demands novel expertise and specialization that (initially) commands a scarcity premium. … If, as we suspect, new work provides additional opportunities for skill formation and earnings growth beyond ‘more work’, then policies that foster new work creation may be of particular interest.
Will rapid advances in Artificial Intelligence shift the balance of innovation towards more rapid automation across an expanding set of occupational domains? What human tasks is AI likely to obviate, what skill sets will AI complement, and what new forms of expertise will become relevant ?
“Early evidence on these questions points to a broad range of potential effects, but definitive findings are elusive so far. … AI innovations are however well-represented in patents, suggesting that the classification methodology developed here may offer insight into AI’s potential for both augmentation and automation.”
Comments