40% of household chores can be automated within 10 years, according to research

A study published last month by researchers from the UK and Japan offered new perspectives on the future of work by including the implications of unpaid work in public discourse. The data showed that time spent on these activities often equaled time spent on paid work.

According to the survey published Feb. 22 in the journal Plos Onerefers to unpaid work, household chores such as cooking, cleaning and laundry, and child and elderly care services. Underscoring that much of the service is disproportionately performed by women, the research question was: “If robots are going to take our jobs, will they at least take out the trash for us?”

The average estimate obtained by the experts showed that 39% of the time currently spent on a household task will be automated within ten years. Interestingly, Japanese male researchers were more pessimistic about this potential, interpreted as a result of gender inequality in Japanese households, showing how these estimates can be socially contingent.

How was the research on the future of unpaid work done?

To prove their hypothesis, 65 British and Japanese researchers conducted a guesswork to estimate how automatizable 17 home and care tasks could be. In addition to assessing a segment not envisioned in previous studies, they also applied a sociological approach that takes into account different biases according to the different origins of the experts.

Research has shown that the degree of automation varies greatly depending on the type of unpaid job involved. Thus, while the chores of caregivers, children or the elderly can be automated at only 28%, the expectation rises to 44% when it comes to strict household chores such as cooking, cleaning and shopping.

Perhaps the biggest contribution of the study is that, whatever the outcome, these predictions are never just technical but also sociotechnical, that is, “predictions about how technology and society will evolve”. Therefore, a technical team evaluating the automation of a task is not empowered to predict its socioeconomic consequences.