As COVID-19 struck the developing world, governments imposed mobility restrictions to reduce contagion, causing severe disruptions in labor markets. In Latin America and the Caribbean, where 57% of all workers hold informal jobs that often require them to be physically present, nearly half of the labor force was compelled to stop work in the first phases of the pandemic.
COVID-19, however, may have changed labor markets in even more profound and lasting ways. Beyond the short-term impacts, the need to limit in-person interactions and reduce contagion may have incentivized firms to replace workers with algorithms, digital devices, or robots. Typically, jobs that involve routine, repetitive, and well-defined tasks are more likely to be replaced by such technology. Moreover, moves towards automation would allow firms to ensure the continuity of production in the face of future pandemics or mobility restrictions.
Filling a Gap in Labor Market Studies of Automation
There is scarce evidence of the implications of COVID-19 restrictions on automation for the developing and developed world. We decided to look at this important question by examining the pandemic’s effect on technology and labor markets in Peru, a highly informal developing economy where the effect of the pandemic was even worse than the average for the Latin American and Caribbean region, with an economic contraction of 11% in 2020 and an initial pandemic phase in which 59% of the labor force was forced to stop working.
Using task descriptions for different types of occupations, we constructed an index of the risk of automation. We relied on individual-level data, which allowed us to track the same workers over time and determine whether they lost their jobs after the pandemic struck. And we leveraged the timing of the COVID-19 shock, policy-induced mobility restrictions imposed on certain industries, and within-industry variation in automation risk to estimate the impact of pandemic restrictions on the move to automated technologies.
The results show that workers at higher risk of this transformation experienced lower employment rates and hourly wages for up to 18 months after the pandemic outbreak. Interestingly, the recovery afterward was not a straightforward rebound, but rather a complex process influenced by a multitude of factors. One of the most significant was occupational mobility. Displaced workers may have adapted to the changing labor market landscape by moving to occupations that had a lower risk of being automated. This suggests that occupational mobility, or the changing of professions, drove the recovery of labor market outcomes for displaced workers.
Automation Especially Affects Certain Groups
Furthermore, the results showed that the negative effects of automation were particularly strong for women, small and medium-sized firms, less-skilled workers, informal workers, and those in the retail, manufacturing, and services sectors. This suggests that pandemic-related mobility restrictions may have helped accelerate technology-adoption by firms that are otherwise tightly constrained in their ability to automate, such as smaller, informal, and service-based ones. These firms likely focused on adopting basic technologies, rather than more sophisticated ones such as robots and artificial intelligence. For example, the use of security cameras to minimize in-person interactions in businesses, such as grocery stores and hair salons, may have reduced the demand for door attendants and other security-related jobs. The implications of these findings are far-reaching. Our study underscores that automation seems to have accelerated in the developing world in recent years. It also shows that workers can adapt to automation through occupational mobility. Policymakers should take note: there is a clear need for training policies that equip workers with cognitive and non-routine skills that can help workers transition to career paths less affected by technology adoption, thereby mitigating automation’s adverse effects.