The truth is not as prophesied

Many had thought that the pandemic would finally vindicate predictions that robots would be job killers. In mid-2020, The National Bureau of Economic Research published a widely cited paper arguing that the coronavirus pandemic “may accelerate the process of job automation”; Another paper asserts that the pandemic “reinforces the trend and impact of automation.” A paper published by the International Monetary Fund also questioned whether jobs lost during the pandemic could be “brought back”. These arguments are based in part on the logic that since robots don’t take sick days, bosses will naturally choose them over workers, as seems to have been the case during previous pandemics. Forecasters stress that intermittent bursts of automation tend to occur during economic downturns.

But two years on, while global investment spending is soaring,330104-00-22-10-02-05  there is little evidence that automation is causing job losses. On the contrary, some developed countries also face the problem of labor shortage. Britain’s The Economist magazine estimates that unfilled job vacancies across the Organisation for Economic Co-operation and Development have reached a record 30 million – contradicting the view that human Labour is no longer needed. Low-skilled jobs are often thought to be more vulnerable to being replaced by robots, yet wages for such workers are growing at an unusually fast rate. In the US, there is still little evidence that highly repetitive jobs are declining relative to other jobs.

What’s more, a new Rutgers University study of the impact of robotics and automation on the economy, jobs and wages in the US from 2009 to 2017 shows that nearly a decade of strong economic recovery has whitewashed the impact of robots on the overall job market, with less-educated workers still not fully displaced.

Statistics show that the number of industrial robots in the 330104-00-22-10-02-05 United States more than tripled between 2009 and 2017, with the ratio of workers to manufacturing robots rising from 0.813 to 1.974 per 1,000 people. From a regional perspective, especially in the Midwest, the robot density is the highest, at least twice that of other regions.

During this same period, the growth of the US job market was significant and steady, and while the use of robots in many factories increased, so did the employment of workers, including young, less-educated men and adult women.

The study argues that while the age of robots is indeed upon us, there is little (or no) evidence that robots are displacing human jobs. However, there are winners and losers in the automation trend, and while the impact of robots on overall employment in the United States is small, it does vary by industry, region and population to some extent.

In addition, the reason why the loss of jobs caused by robots has little impact on the overall employment rate may be that in an environment of strong economic growth and low unemployment, even if workers are displaced, they can still find other jobs, but they may pay less.

Automation may actually increase employment

Recently, economists Philippe Agion, Schilling Antonin, Simon Binel, and Xavier Jarravel from research institutions in France and the United Kingdom have proposed a new idea about robots – that the direct result of automation may be to increase employment at the firm level, rather than reduce it. This sounds counterintuitive, but there is a solid microeconomic basis for it: automation is likely to make companies more profitable, so they 330104-00-22-10-02-05 will expand and hire aggressively. At the same time, technological innovation may also allow companies to further explore new areas or focus on more labor-intensive products and services.

According to The Economist, a growing body of research supports this view. Researchers at Yale University in the US looked back at Japanese manufacturing from 1978 to 2017 and found that each additional robot per 1,000 workers increased the number of employees by 2.2 per cent.

Researchers at the Massachusetts Institute of Technology in the United States took a closer look at Finnish companies and found that adopting advanced technology actually led to hiring more workers. In an as-yet-unpublished study, researchers from Stanford University in the US and the London School of Economics in the UK looked at the use of machine tools in UK industry and found that automation “has a strong positive correlation with firm survival, and that higher levels of initial automation are associated with increased employment”.

While economists may now appear to have performed a 330104-00-22-10-02-05 U-turn, they cannot simply be said to have been wrong.

Statistical methods have improved a lot since the paper on roboeconomics was published. These papers, published in 2013 by Carl Benedikt Frey and Michael Osborne of the University of Oxford in the United Kingdom, were widely interpreted as suggesting that 47 percent of the U.S. workforce was at risk of being displaced by automation. In fact, one of the difficulties of research is untangling cause and effect: companies that hire heavily may also buy robots, rather than the other way around. But their paper shows that companies buy robots when they become cheaper. This creates a causal chain: cheaper robots lead to more automation, which leads to more employment.

It is also important to note that the New View does not assert that automation is good. So far, it has not revealed how the quality of jobs affects workers’ wages. But David Autor, David Mindell and Elizabeth Reynolds of the Massachusetts Institute of Technology have found that even if the proliferation of robots does not cause widespread unemployment, it may have helped create an environment in which the reward system is “tilted upwards”. Others argue that automation degrades the quality of jobs.

MIT’s Mr Agion adds that even if automation boosts employment at the company or industry level, its impact on the wider economy is less clear. In theory, businesses that adopt robots could be so successful that they might even drive competitors out of business, reducing the total number of jobs. These questions leave researchers with much wider scope to explore. But what seems clear at this stage is that the prevailing pessimism about automation is well and truly over.