W.Hen the When the epidemic hits first, unemployment rises. Since the depression American unemployment has exceeded 14%, as happened in April 2020. But long-term fears of high unemployment have not been met. According to the latest available data, the unemployment rate for November OECD Clubs in most rich countries were slightly higher than before the epidemic. So far it may even draw layers. Bouncebacks in the labor market of the rich world are the latest developments that have prompted economists to reconsider a fundamental question of discipline: do robots help or harm workers.
The melancholy narrative, which says that the attacks of job-killing robots are almost at an angle, has held a tremendous hold on popular imagination for decades. The warnings about the future of the unemployed have created ridiculously enough, plenty of jobs for ambitious public intellectuals who are looking for book deals or talking opportunities. Shortly before the epidemic, however, other researchers began to question the knowledge gained. The world was supposed to be in the middle of an artificial intelligence and machine-learning revolution, but by 2019 the employment rate in developed economies had reached an all-time high. Japan and South Korea, where robot use was highest, had the lowest unemployment rates.
Many thought that the epidemic would eventually prove to be a catastrophe. A highly cited paper published by the U.S. National Bureau of Economic Research in mid-2020 argued that Covid-19 could “accelerate job automation”, and another stressed that it was “strengthening the trend of both automation and its impact.” A research paper published by IMF Wondered if the job lost during the epidemic would “come back”. Part of the argument was that since robots don’t get sick, Basra would return to them instead of humans – as had been the case with some previous epidemics. Others have noted that explosions of automation tend to occur during recessions.
Although two years later, evidence of autoimmune-induced unemployment is very low, even global investment costs are rising. Wealthy world faces labor shortage – we have a record 30m unfinished vacancies worldwide OECDIt’s hard to reconcile this notion with what people no longer need. Wage increases for low-skilled workers, whose occupations are generally considered more risky to be replaced by robots, are unusually fast. There is still very little evidence from America that “routine” jobs, which are automatically considered easy, are shrinking compared to other types of work.
Considering that a lot of skepticism has arisen about the “robot kills job” narrative, it is not surprising that a different thesis is emerging. In a recent study, Philip Aghion, Celine Antonin, Simon Bunnell and Xavier Zaravel, economists from a range of French and British institutions, offer a “new perspective” on robots, saying that “the direct effect of automation may be to firmly increase employment, not reduce it.” ” This view, which may sound pagan, has a strong microeconomic basis. Automation can help a firm become more profitable and thus expand, creating a hiring trend. Technology may allow companies to move to new areas, or focus on more labor-intensive products and services.
A growing body of research supports the argument. Daisuke Adachi and colleagues at Yale University looked at Japanese production from 1978 to 2017. They found that an increase in robotic units per 1,000 workers increased the firm’s employment by 2.2%. Another study by Junas Tuhkuri of the Massachusetts Institute of Technology (With) And colleagues, look at Finnish firms and conclude that their adoption of advanced technology has resulted in increased employment. The unpublished work of Michael Webb of Stanford University and Daniel Chandler of the London School of Economics examines machine tools in British industry and discovers that automation “had a strong positive relationship with survival, and was associated with greater early automation employment growth”.
Non-economists can be forgiven for turning a blind eye to the apparent appearance of their profession. But things are not so simple that it is said that economists have done it wrong before. For a start, the statistical approach has improved since the publication of basic research papers on robonics, such as those of Carl Benedict Frey and Michael Osborne at Oxford University in 2013, which widely explained that 47% of Americans were at work risk. Automation The method used by Mr. Adachi and his co-authors is particularly clever. One problem is persistence of causation: Recruitment firms may be able to buy robots, but not otherwise. But the paper shows that companies buy robots when their prices go down. It helps to establish a causal chain from cheaper robots, to more automation, to more jobs.
Strong wave of research …
A second qualification is that the “new approach” does not establish that automation is “good”. So far, it can say very little about the quality of work and wages. But an upcoming book by David Autor, David Mindell and Elizabeth Reynolds With Shows that while robots don’t create massive unemployment, they can help create an environment where rewards are “slanted to the top.” Others argue that automation reduces the quality of work.
Mr Aghion and his colleagues add that even if automation increases employment at the firm or industry level, its impact on the economy as a whole is less obvious. Theoretically robot-adoption companies can be so successful that they drive competitors out of business, reducing the total number of jobs available. Such questions leave much more for researchers to explore. But what seems clear at this point is that the era of sweeping, depressing narratives about automation is well and truly over. 3
Read more from Free Exchange, our column on economics:
Will remote work be stuck after the epidemic? (January 15)
IMF reprimands IMF over Argentina (January 8)
New research calculates cost of Sino-American trade war (January 1)
For more expert analysis of the biggest stories in economics, business and markets, sign up for our weekly newsletter Money Talks.
This article was published in the Finance and Economics section of the print edition under the heading “Updates in Progress”