Risks of automation of occupations: An estimate for Italy

There is a rising concern for technological unemployment due to the current digital revolution. In this paper, we estimate the probability of automation of the 800 Italian occupations, evaluate the distribution of Italian workers according to their risk of substitution, and suggest some strategies that could be adopted to reduce that risk. On the one hand, Italy ranks second in Europe for robot stock. On the other hand, the adoption of new technologies by firms is uneven and Italian firms lag behind in the adoption of new production technologies. To estimate the probability of automation of Italian occupations we applied two approaches: the occupation-based one (Frey and Osborne 2017) and the task-based one (Nedelkoska and Quintini 2018). In line with previous literature, the probability of automation obtained applying the task-based approach are in general lower than those estimated using the other approach. For most occupations the probability of automation under the two approaches differs by about 10-20%, although for some the difference is more significant. Italian occupations with a high probability of automation require many routine activities, whereas those with a low probability of automation require abilities like perception, manipulation, creative intelligence and social intelligence. Based on the occupation-based approach, 33.2% of Italian workers face a high risk of replacement; this percentage decreases at 18.1% if we apply the task-based approach. Male workers appear to face a higher risk of replacement than female ones regardless of the approach adopted. Based on our results and considering the characteristics of the Italian context, we discussed the factors that impact on real automation and suggested some policies that could be adopted