July 24, 2017
Is the Robocalypse upon us?, asked MIT economist David Autor in his presentation at a recent Forum of European central bankers. His presentation was based on a paper co-written with Utrecht University economist Anna Salomons. “Is productivity growth inimical to employment?,” they asked in the paper’s abstract. “Canonical economic theory says no, but much recent economic theory says maybe - that is, rapid advances in machine capabilities may curtail aggregate labor demand as technology increasingly encroaches on human job tasks.”
Fears that machines will put humans out of work are not new. Throughout the Industrial Revolution there were periodic panics about the impact of automation on jobs, going back to the Luddites, - textile workers who in the 1810s smashed the new machines that were threatening their jobs. But, “In the end, the fears of the Luddites that machinery would impoverish workers were not realized, and the main reason is well understood,” noted a 2015 article on the history of technological anxiety.
“The mechanization of the early 19th century could only replace a limited number of human activities. At the same time, technological change increased the demand for other types of labor that were complementary to the capital goods embodied in the new technologies. This increased demand for labor included such obvious jobs as mechanics to fix the new machines, but it extended to jobs for supervisors to oversee the new factory system and accountants to manage enterprises operating on an unprecedented scale. More importantly, technological progress also took the form of product innovation, and thus created entirely new sectors for the economy, a development that was essentially missed in the discussions of economists of this time.”
Automation fears have understandbly accelerated in recent years, as our increasingly smart machines are now being applied to activities requiring intelligence and cognitive capabilities that not long ago were viewed as the exclusive domain of humans. “Previous technological innovation has always delivered more long-run employment, not less. But things can change,” said a 2014 Economist article. “Nowadays, the majority of economists confidently wave such worries away… Yet some now fear that a new era of automation enabled by ever more powerful and capable computers could work out differently.”
To shed light on these automation fears, Autor and Salomons explored the relationship between productivity growth and employment by analyzing country- and industry-level data collected from 1970 to 2007 The data comes from 19 rich countries including the US, UK, Japan, Germany and France; and from 28 industries, which they combined into five mutually exclusive sectors: manufacturing; mining, utilities and construction; education and health services; capital intensive high-tech services; and labor intensive low-tech services.
Their paper addressed three major questions: Has productivity growth threatened employment, especially in advanced economies? What’s the impact of productivity growth on employment at the industry level? And, are there productivity implications for labor demand by skill group? Details of their analysis can be found in the paper and presentation. Let me attempt to summarize their key findings.
Has productivity growth threatened employment?
Their overriding conclusion is that “over the 35+ years of data that we study, we find that productivity growth has been employment-augmenting rather than employment-reducing; that is, it has not threatened employment… This strong finding emerges despite robust evidence that industries experiencing rising labor productivity exhibit falling employment…”
“The reason that industry-level productivity growth typically raises net employment is because productivity growth in each sector - particularly in services - generates employment growth spillovers elsewhere in the economy. These spillovers are sufficiently large that they more than offset employment losses in industries making rapid productivity gains. Individually, we estimate that both the employment-reducing and employment-increasing effects of productivity growth are economically sizable: however, their net effect makes for a rather modest positive impact of productivity growth on employment…”
“Perhaps the most surprising - and yet simultaneously, pedestrian - takeaway from our analysis is that productivity growth is not the primary driver of rising or falling employment. We estimate that net employment changes resulting directly or indirectly from productivity growth are quite modest, amounting to only a few percentage points of net employment over more than three decades. Instead, the primary driver of employment growth is estimated to be population growth; the number of workers rises roughly in lock-step with the overall growth of citizens in a country. This observation, which is almost self-evident but not tautological, suggests that the conventional narrative in which automation is the critical factor in either eroding or augmenting employment misses the mark.”
Impact on industry sectors
Economic activity is generally organized into three distinct sectors: the primary sector comprises industries dealing with natural resources such as agriculture, mining, fishing; the secondary sector includes industries that produce manufactured goods as well as construction; and the tertiary sector includes industries that provide services. The rapid productivity growth in primary and secondary industries throughout the 20th century has led to a substantial reallocation of jobs into the tertiary or service sector.
“Over the comparatively short timespan between 1970 – 2007, the share of employment in manufacturing dropped by more than 15 percentage points while the share in mining, construction, and utilities fell by roughly three percentage points. Conversely, the share of employment in education and health rose by eight percentage points, the share in high-tech services rose by ten percentage points, and the share in low-tech services rose by a modest two percentage points.”
The impact of productivity growth on aggregate employment varies considerably across industry sectors, with productivity growth in services showing the most impact. “Given that service sector productivity growth appears to have relatively strong employment spillovers, our findings suggest that the productivity growth spurred by the spread of (ultimately) general-purpose technologies such as robotics from heavy industry and into services may prove a boon for employment growth.”
Implications on different skill groups
Over the past 15 years, professor Autor has written extensively on the polarization of job opportunities by skill. While his research has been primarily focused on the US labor market, the findings are generally applicable to other advanced economies. His research has uncovered the changing dynamics and job opportunities across different skill groups:
High skill: These jobs generally require the kinds of expert problem solving and complex communications skills typically seen in managerial, professional and technical occupations. Over the past 30 years, high-skill jobs have significantly expanded, with the earnings of the college educated workers needed to fill such jobs rising steadily.
Mid skill: Many of these jobs deal with the kinds of routine tasks that can be well described by a set of rules and have thus been prime candidates for automation. Many blue collar jobs, such as manufacturing and other forms of production, fall into this category. So do white-collar, information-based activities like accounting, record keeping, and different kinds of administrative tasks. Mid-skill jobs have been steadily declining, especially since 2000.
Low skill: These jobs generally involve physical tasks that cannot be well described by a set of rules and are therefore not easily automated. They include protective services, food and cleaning services, personal care and health care aides. Low-skill jobs have been expanding, while their wage growth, particularly since 2000, has been flat to negative.
Job opportunities have sharply polarized over the past few decades, with expanding opportunities in both high- and low-skills occupations and contracting opportunities in mid-skill occupations. This is particularly true for services, - the fastest growing job segment in the economy.
“Specifically, these productivity-induced sectoral shifts are shown to be sharply biased in favor of skilled workers. In this respect, our analysis underscores a central insight of much recent work on the labor market impacts of technological progress: the primary societal challenge posed so far by these advances is not falling aggregate labor demand but, rather, an increasingly skewed distribution of employment - and ultimately earnings - favoring highly educated workers.”
Finally, as every investment prospectus reminds us, past performance is no guarantee of future result. The AI and robotics revolution is just getting going. It will clearly have a major impact on jobs and the very nature of work, but it’s much less clear what that impact will be. Will it, once more result in more long-run employment? Or, will this time be different? Opinions abound, but in the end, we don’t really know.