Technology and work: Technological unemployment and will this time be different?
Phuah Eng Chye (5 May 2018)
The history of technological innovation is the history of social dislocation. The transition from agriculture to manufacturing was notable for its massive impact on employment and societal structures. The effect of technology on society still occupies the centre of much economic debate; often drawing their roots from a centuries-old ideological feud between Marxism and capitalism.
The recent debate on technological unemployment resurfaced after the global financial crisis of 2007. High-profile studies drew attention to how the speed and extent of technology in replacing human workers is accelerating; carrying with it the potential threat of a job Armageddon. In the meantime, dissatisfaction among the population became widespread and spilled over into election outcomes.
My analysis of the debate is spread over a series of articles on technology and work, the labour share of income and the future of work. This article starts by examining the basic proposition of whether technology will destroy or create more jobs; i.e. the effect of technology on the level of unemployment.
Martin Ford suggests a big shift is occurring due to the new technologies based on robotics, artificial intelligence, big data, advanced self-service automation and 3D printing. The new technologies elevate the threat of job elimination from automating repetitive jobs to replacing predictive jobs; threatening professions such as “lawyers, journalists, scientists and pharmacists”.
In addition, “technology can leverage the efforts of a tiny workforce into enormous investment value and revenue…they offer compelling evidence of how the relationship between technology and employment has changed…As a result, emerging industries will rarely, if ever, be highly labor-intensive.”
Martin Ford expects “the threat to employment is that as creative destruction unfolds, the destruction will fall primarily on labor-intensive businesses in traditional areas like retail and food preparation, while the creation will generate new businesses and industries that simply don’t hire many people. In other words, the economy is likely on a path toward a tipping point where job creation will begin to fall consistently short of what is required to fully employ the workforce.”
Supporting this dire scenario, a widely-cited 2013 study by Carl Benedikt Frey and Micahel A. Osbourne estimated 47 percent of total US employment was at risk with transportation, logistics, office and administrative support, and some manufacturing and production jobs the most likely to be eliminated while service occupations such as food service and retail jobs were the most vulnerable to automation.
The validity of this study has been disputed. Robert D. Atkinson and John Wu criticised the Osborne and Frey study; noting the study was not submitted for “peer review, neglected to examine all 702 US occupational categories to manually assess how likely it is that technology will substitute for a human worker…their methodology produces results that make little sense, as when they predict that technologies such as robots will eliminate the jobs of fashion models, manicurists, carpet installers, and barbers…When the Information Technology and Innovation Foundation (ITIF) analyzed these 702 occupations manually, using a very generous assumption about how tech could eliminate jobs, we estimated that about 10 percent of jobs were at risk of automation, at most.”
Other studies also suggest a more subdued impact. Daron Acemoglu and Pascual Restrepo studied the impact of industrial robot usage on US local labor markets between 1990 and 2007. They estimated “one more robot per thousand workers reduces the employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent.” They add “so far, there are relatively few robots in the US economy, and so the number of jobs lost due to robots has been limited to between 360,000 and 670,000 jobs…it should also be noted that even under the most aggressive scenario, we are talking about a relatively small fraction of employment in the US economy being affected by robots.”
Ljubica Nedelkoska and Glenda Quintini notes while some initial reports “produced estimates in the high double digits” but recent studies “brought the estimates of the share of jobs at risk of automation down significantly” based on the “considerable variation in the tasks involved in jobs having the same occupational title.”
Generally, optimists base their case on technology’s track record in creating more jobs than they destroy. David H. Autor notes “despite sustained increases in material standards of living, fear of the adverse employment consequences of technological advancement has recurred repeatedly in the 20th century…anxiety about the adverse effects of technological change on employment has a venerable history. In the early 19th century, a group of English textile artisans calling themselves the Luddites staged a machine-trashing rebellion in protest of the rapid automation of textile production, which they feared jeopardized their livelihoods.” He points out these fears “overstate the extent of machine substitution for human labor and ignore the strong complementarities that increase productivity, raise earnings, and augment demand for skilled labor.”
Hence, optimists expect technological innovation will continue to spawn new industries and lifestyles and create jobs not previously envisaged. Robert D. Atkinson and John Wu note “in 2012, there were 466,000 U.S. jobs related to mobile apps, up from zero in 2007…examine the fastest-growing U.S. industries over the last 15 years, certainly some are due to technological innovation…support activities for oil and gas operations grew by 537 percent, in part to support natural-gas fracking…Many fast-growing industries are, not surprisingly, in the IT industry, such as internet publishing, internet services providers, software, and cellular communications systems. Others – such as biological products and surgical and medical instrument manufacturing – are also spurred by innovation, enabling new products to come to market.”
Robert D. Atkinson and John Wu also point out the “levels of occupational churn in the United States had reached historic lows. Their analysis showed “occupational churn peaked at over 50 percent in the two decades from 1850 to 1870 (meaning the absolute value sum of jobs in occupations growing and occupations declining was greater than half of total employment at the beginning of the decade), and it fell to its lowest levels in the last 15 years – to around just 10 percent. When looking only at absolute job losses in occupations, again the last 15 years have been comparatively tranquil, with just 70 percent as many losses as in the first half of the 20th century, and a bit more than half as many as in the 1960s, 1970s, and 1990s. Hence, “in contrast to the popular view that technology today is destroying more jobs than ever, our findings suggest that is not the case. The period from 2010 to 2015 saw approximately 6 technology-related jobs created for every 10 lost, which was the highest ratio – meaning lowest share of jobs lost to technology – of any period since 1950 to 1960.”
In its November 2017 report, the McKinsey Global Institute (MGI) estimated “while about half of all work activities globally have the technical potential to be automated by adapting currently demonstrated technologies, the proportion of work actually displaced by 2030 will likely be lower, because of technical, economic, and social factors that affect adoption. Our scenarios across 46 countries suggest that between almost zero and one third of work activities could be displaced by 2030, with a midpoint of 15 percent.”
The MGI report also reviewed the this time is different arguments. It examined data to gauge whether the pace of automation innovation was faster, its effects more pervasive across sectors, and affected workers at a broad range of skill and wage levels. After comparing data from past waves of technology disruption, they concluded “in many respects, the impact of automation on employment today is not likely to be different than in the past…no evidence that technological adoption has yet accelerated over the last 60 years…little is new about the breadth of impact of automation technologies…today’s automation is unlikely to be different from the past”.
They note “if history is any guide, we could expect 8 to 9 percent of 2030 labor demand will be in new types of occupations that have not existed before.” In addition, job change will be a major feature of the transition. “Our scenarios suggest that by 2030, 75 million to 375 million workers (3 to 14 percent of the global workforce) will need to switch occupational categories.”
Karen Harris, Austin Kimson and Andrew Schwedel suggests the dislocation effects vary depending on the speed of rollout of automation. “If automation rolls out slowly, workers who lose their jobs will have more time to adjust, retrain or simply retire out of the workforce. A slow pace would avoid an abrupt spike in investments for newly automated capacity and create a relatively gentle push and pull between the need for more output capacity (due to unfavorable demographics) and more output potential (due to increased automation). In this scenario, the world economy would muddle through the shift to a more automated economy with lackluster growth but also less disruption.” But they note “a slow transition flies in the face of data on the quickening pace of technological adoption over the past half century.”
“But rapid automation of the US service sector, for example, could eliminate jobs two to three times more rapidly than in previous transformations. As the investment wave recedes, it may leave in its wake deeply unbalanced economies in which income is concentrated among those most likely to save and invest, not consume. Growth at that point would become deeply demand constrained, exposing the full magnitude of labor market disruption temporarily hidden from view by the investment boom. Over time, the interplay of these three major forces may produce an apparent contradiction—a dramatic surge in output potential that ultimately leads to stagnation”.
Karen Harris, Austin Kimson and Andrew Schwedel observe “while the pace of technological change is arguably accelerating, we have seen no evidence yet that the rate of human adaptation to jarring economic dislocations has improved. If anything, the experience of the two recent US recessions points in the opposite direction – an aging labor force is becoming less able to learn new skills and find work. The demographic outlook for the next decades suggests that the labor force’s speed of adjustment to disruptions might actually worsen. Our analysis suggests that the pace of labor force displacement in the coming decade could be two to three times as fast as during other big transformational periods of labor automation in modern history.”
Hence, they “expect the magnitude of workforce change in the 2020s to match that of the automation of agriculture from 1900 to 1940. However, the transition of farm workers into the industrial sector was spread out over four decades. In the case of the automation of manufacturing, the impact was over a shorter time period (roughly 20 years), but the share of labor force in manufacturing jobs was relatively small in the US. Investment in automation is likely to proceed moderately faster than agricultural automation or manufacturing automation unless other forces act to impede its progress, and it will affect a larger percentage of the total workforce”.
Ljubica Nedelkoska and Glenda Quintini caution that even if job losses are smaller in magnitude than estimated, since the “job losses are unlikely to be distributed equally across the country, this would amount to several times the disruption in local economies caused by the 1950s decline of the car industry in Detroit where changes in technology and increased automation, among other factors, caused massive job losses.” In particular, “the regional concentration of the risk of automation could amplify its social and economic impact, particularly in countries where geographical mobility is low” while certain groups such as the younger age groups and lower-skilled workers will be disproportionately affected”.
Morgan Frank, Manuel Cebrian, Hyejin Youn, Lijun Sun and Iyad Rahwan found job losses will generally be higher in small cities than large cities. “While cities of all sizes have many easily automated jobs (like card dealers, fisherman, cashiers, and accountants), large cities like Boston also have larger shares of managerial and knowledge professions (like lawyers, scientists, and software developers). Since these jobs require knowledge and skills that cannot easily be taught to a machine, they will offset the total impact of automation. In smaller cities, fewer of those offsetting jobs exist. The “more intense job loss will push people to leave smaller cities” to find work in a larger city.
Overall, there is discomfort due to the lack of visibility on the prospects for new job creation. This creates doubts as to whether sufficient jobs can be created to absorb the displaced workers. But I suspect the employment uncertainties are being overshadowed by the difficulty in extracting higher productivity from technology and by the way technology is changing the nature of work. In relation to the latter, quantitative comparisons are more meaningful when employment is discrete – employed or not-employed. They become less meaningful as jobs fragment and become more transient. Permutations in the types of employment and changes in the nature of work are likely to pose a greater challenge that the availability of jobs.
Carl Benedikt Frey, Michael A. Osbourne (17 September 2013) “The future of employment: How susceptible are jobs to computerization” Oxford Martin School Programme on the impacts of future technology Working Paper. http://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf
Daron Acemoglu, Pascual Restrepo (March 2017) “Robots and jobs: Evidence from US labor markets”. NBER Working Paper No. 23285. http://www.nber.org/papers/w23285.pdf
David H. Autor (3 September 2014) “Polanyi’s paradox and the shape of employment growth”. NBER Working Paper No. 18629. www.kc.frb.org/publicat/sympos/2014/2014093014.pdf
James Manyika, Susan Lund, Michael Chui, Jacques Bughin, Jonathan Woetzel, Parul Batra, Ryan Ko, Saurabh Sanghvi (December 2017) “Jobs lost, jobs gained: Workforce transitions in a time of automation.” McKinsey Global Institute. file:///C:/Users/user/Downloads/MGI%20Jobs%20Lost-Jobs%20Gained_Report_December%202017%20(1).pdf
Karen Harris, Austin Kimson, Andrew Schwedel (7 February 2018) “Labor 2030: The collision of demographics, automation and inequality”. Bain & Company. http://www.bain.com/publications/articles/labor-2030-the-collision-of-demographics-automation-and-inequality.aspx?
Ljubica Nedelkoska, Glenda Quintini (14 March 2018) “Automation, skills use and training”. Directorate for Employment, Labour and Social Affairs. OECD. https://www.oecd-ilibrary.org/docserver/2e2f4eea-en.pdf?expires=1522802514&id=id&accname=guest&checksum=CAA690E07BBE09C9D0B781D2122B98DF
Martin Ford (2015) The rise of the robots: Technology and the threat of mass unemployment. Oneworld Publication.
Morgan Frank, Manuel Cebrian, Hyejin Youn, Lijun Sun, Iyad Rahwan (10 April 2018) “How will automation affect different U.S. cities?” Kellog Insight. https://insight.kellogg.northwestern.edu/article/how-will-automation-affect-different-united-states-cities
Robert D. Atkinson, John Wu (8 May 2017) “False alarmism: Technological disruption and the U.S. labor market, 1850–2015.” The Information Technology and Innovation Foundation (ITIF). https://itif.org/publications/2017/05/08/false-alarmism-technological-disruption-and-us-labor-market-1850-2015