Recent Research: Automation not resulting in greater job loss at the country level

Discussions surrounding automation’s power and the effect it could have on jobs have only increased over time. The current pandemic adds to the debate of whether automation and robotics, which are unaffected by viruses and have the potential for cost savings, could offer a safer bet for industries than human labor. Such are the debates the authors of a new working paper considered in their research examining jobs that were identified in the past as being at risk of elimination through automation. While building on previous studies from the Organisation for Economic Cooperation and Development (OECD) of the impact of automation on jobs, OECD authors Alexandre Georgieff and Anna Milanez seek to expand that knowledge to a cross-country context and the paper claims to be the first to evaluate employment outcomes using the task-based measure of automation risk developed by the OECD. The researchers found no support for net job destruction at the broad country level. However, they did find evidence that automation has worsened employment prospects for some workers including skilled agricultural workers, clerical support workers, and metal and machinery workers.      

In "What happened to jobs at high risk of automation?", OECD Social, Employment and Migration Working Papers, the authors set about to learn whether countries and jobs that were said to be at high risk of automation in 2012 actually experienced employment declines over the subsequent period, whether certain demographic groups were more vulnerable to those changes and how job stability has changed in those occupations. What they found was that countries that faced higher automation risk in 2012 experienced higher employment growth over the subsequent period of 2012 to 2019. They also found that the majority of occupations saw employment growth, but those with higher risk of automation had lower employment growth or even saw declines, as mentioned above.

Although low-educated workers were not more negatively affected by automation trends, the authors note that job opportunities for these workers were accompanied by a decline in the number of low-educated workers as the result of the general upskilling of the workforce. However, low-educated workers have become increasingly concentrated in high-risk occupations.

The authors conclude that “anxiety around a jobless future may be overblown.” But they also recommend that policy should focus on attention on those who may not share evenly in the benefits of automation, namely, the low-educated and they note that the “potential for automation to produce worse outcomes among occupations at high risk of automation is even more worrisome in the present context of the severe employment declines brought by COVID-19.” They suggest training programs to accompany individuals through job transitions, social protections to cushion workers and education programs to prepare young workers for new, higher-skilled jobs.

The OECD working paper is available here.