The impacts of digital transformation on the labour market: Substitution potentials of occupations in Germany

The digital transformation will have large effects on the labour market. Previous studies mostly reveal that half of all current jobs are susceptible to automation in the next 10 to 20 years. These studies use assessments by technology experts on the future automation probabilities of occupations. We, instead, assume that only certain tasks in an occupation, rather than entire occupations, can be substituted. We directly calculate automation probabilities – labelled as substitution potentials – for occupations in Germany by using German occupational data from an expert database. Considering approximately 8000 tasks, we assess whether they can be replaced by computers or computer-controlled machines according to programmable rules. We do not give a forecast for the future but estimate existing technological possibilities. We argue that previous studies overestimate automation probabilities because they do not start with the tasks. We demonstrate that we obtain values similar to those in previous studies when assuming that entire occupations are replaceable: approximately 47% of German employees work in a substitutable occupation in 2013. Assuming that only certain tasks can be substituted, we find that only 15% of German employees are at risk. Furthermore, we provide evidence on the relationship between automation risks and employment growth. •We determine the impacts of digital transformation by a task-based approach.•Almost any occupation includes tasks that are not substitutable.•Occupation-level approaches overestimate automation probabilities.•Using a task-based approach, only 15% of jobs are at risk of being automated.•Employment is growing less rapidly with increasing substitution potentials.