Automation of job tasks may affect women more than men
Recent advances in artificial intelligence have raised concerns about the potential impact of automation in the workplace. The COVID-19 pandemic may accelerate the implementation of new technology, as firms might look to make the production and delivery of goods and the provision of services more resilient in the future. While skilled workers may become more productive by complementing the tasks performed by the new technology or by working directly with it, others may need to upgrade their skills. In either case, jobs may be transformed, as robots and computer algorithms take over routine, non-cognitive duties, while humans specialize further in non-routine, cognitive tasks. These changes may affect women and men differently, depending on the tasks they perform and how automatable they are.
A new Statistics Canada study is the first to examine in detail the risk of automation-related job transformation faced by women and men based on their job tasks.
The study finds that in 2016 (prior to COVID-19), women and men were equally likely to face a high risk of automation-related job transformation (about 11%). However, women (44.4%) were more likely than men (34.8%) to face a moderate to high risk.
The gap in the proportion of women and men facing a moderate to high risk of automation-related job transformation could not be explained by gender differences in personal and work characteristics, such as age, education, industry and occupation. The gap may indicate that women and men perform different tasks that are not taken into account in the data. In fact, previous research has shown that women were more likely than men in the same occupation to report performing repetitive tasks, and this could put them at greater risk of automation-related job transformation.
The higher share of women facing a moderate to high risk of automation-related job transformation, compared with men, was also observed within many subgroups of the population. In certain cases, the gap was particularly large. For example, while 33.9% of men aged 55 or older faced a moderate to high risk, 58.6% of their female counterparts faced a similar risk. A larger share of women with no postsecondary qualifications faced a moderate to high risk than their male counterparts (75.8% and 60.0%, respectively).
Women who reported having a disability, who were not in a union or covered by a collective bargaining agreement, or who worked in a small firm (with 10 or fewer employees) were also more likely than their male counterparts to face a moderate to high risk of automation-related job transformation.
It is important to note that these risk estimates are largely based on the technological feasibility of automating job tasks. There are several reasons why employers may not immediately replace humans with robots, even if it is technologically feasible to do so. These reasons include financial, legal and demand-side factors, among others. For this reason, a high risk of automation does not necessarily imply a high risk of job loss. That being said, these results were estimated prior to COVID-19, which may accelerate automation in the workplace.
Note to readers
The study uses the 2016 Longitudinal and International Study of Adults and draws on previous research that provides estimates of automation risk by occupation. Automation risk estimates are produced by various worker and firm characteristics and account for 25 different tasks that may vary within the same occupation, such as instructing, selling products or services, solving problems, or performing physical work. A high risk of automation-related job transformation is defined as a probability of 70% or more, while a moderate to high risk is defined as a probability of 50% or more.
The study Automation and the Sexes: Is Job Transformation More Likely Among Women?, part of the Analytical Studies Branch Research Paper Series (11F0019M), is now available.
For more information, contact us (toll-free 1-800-263-1136; 514-283-8300; STATCAN.infostats-infostats.STATCAN@canada.ca).
To enquire about the concepts, methods or data quality of this release, contact Marc Frenette, 613-864-0762; firstname.lastname@example.org, Social Analysis and Modelling Division.