Tuesday, 30 November 2021

Jobs: when machines compete with humans

A study involving UniTrento DEM makes an estimate of jobs that may be lost to automation

Versione stampabile

People employed in accounting, deliveries, shop checkouts, phone services, front desks and assembly lines are among the those who are most exposed to the risk of losing their jobs to automation. On the other hand, entrepreneurs and people employed in child care, care homes and personal assistance services have nothing to fear: their skills in complexity management, empathetic interpersonal relationships and creative intelligence make it very difficult to replace them.

These are some of the findings of the study "Rischi di automazione delle occupazioni: una stima per l’Italia" (risk of job losses from automation: an estimate for Italy), published in issue no. 3 of the journal "Stato e Mercato" by Mariasole Bannò (formerly the University of Trento, now the University of Brescia) with Emilia Filippi and Sandro Trento (University of Trento). The work is part of the analysis of digital technologies in the world of work and of the impact they have on the working population in terms of number of affected workers and inequalities for the profiles that can easily be replaced. For the first time, Bannò, Filippi and Trento have accurately calculated the probability of automation for 800 jobs in Italy and offer suggestions to mitigate the risks.

"We considered three technical limitations in our estimate of the likelihood of automation. These limitations are connected to abilities that only humans possess: the ability to perceive and manipulate objects (including in complex situations and with no context), creative intelligence (the ability to generate new, useful ideas), and social intelligence (the ability to interact with humans with cleverness and empathy)", the authors explained.

We followed a dual approach to study the Italian case: an occupation-based approach, based on the idea that whole occupations are automated, and a task-based approach, which considers single tasks within occupations. According to the type of approach, the percentage of workers at high risk of technological replacement in the coming years in Italy varies between 33% (7.12 million people) and 18% (3.87 million).

In Italy, jobs that are most at risk of automation are in transportation and logistics, administration, industry, services and the sales sector. On the other hand, jobs with a low likelihood of automation are in management and finance, the legal sector, education, healthcare, art. "Jobs at low risk of automation require a high level of education and include a significant share of tasks that can only be performed by humans with creativity, adaptation and interpersonal skills, training, the ability to collaborate with other people", they underline. The manufacturing and media sectors are in the middle.

They continued: "We have also found out that the likelihood of automation is different for male and female workers. Men's jobs are more likely to be affected by job automation. The reason for this could be the different job distribution by gender. In particular, in Italy female employment is higher in sectors with less automation (for example in early childhood education, personal services, healthcare, the food and agriculture industry)".

The study then focuses on the discrepancy between potential automation (level of automation that is technically possible) and actual automation (the level of automation that is actually achieved). The Italian economy is characterized by many small and medium-sized enterprises, often with family-run, a reduced investment capacity, the low adoption of advanced technologies and a series of other factors, for which the actual automation could be less than expected. But, apart from this, the loss of jobs and the decline of some professions represent a danger for Italy and that is why, at the end of the study, the authors offer some advice.

First, new positions with non-automatable tasks must be created to employ those who will lose their job to automation and protect them from the risk of further replacement. Jobs must therefore be created in the personal services sector, tourism, healthcare and education. "To this end, it may be necessary to develop funding for innovation to support the most dynamic start-ups and reduce labor tax and tax wedge to make human work profitable".

The second action involves education and training before entering the world of work and lifelong learning. "In addition to technical skills, creative skills, problem solving abilities, and social and relational skills seem to be important. Education should therefore be redesigned to include school-work relationships and apprenticeships and provide upskilling and reskilling for adults".

Finally, Bannò, Filippi and Trento focus on the rapid increase of job automation is Italy: "The speed at which new technologies seem to replace jobs makes it necessary to reflect on income support policies for the unemployed".