Academic literature suggests that technological advances have caused employment polarization in European Union countries over the past few decades. Computer technology and robots have to some extent replaced routine, medium-skilled jobs such as machine operations, construction work, and administrative tasks, but they have also led to an increase in complementary, non-routine, high-skilled jobs (e.g., managers, specialists) and low-skilled jobs (e.g., agriculture, cleaning, personal care services). But our new research suggests that new technologies that have emerged since 2010, artificial intelligence and machine learning, will profoundly change the landscape of work in the coming decades. These technologies could have a deeper impact on a wider range of jobs and tasks, including the potential for the disruption of low-skill jobs.
(…) New technologies that have emerged since 2010, artificial intelligence and machine learning, will profoundly change the landscape of work in the coming decades. These technologies could have a deeper impact on a wider range of jobs and tasks, including the potential for the disruption of low-skill jobs.
Artificial intelligence (AI) systems can perform tasks that involve decision-making, changing the impact of automation on the workforce. AI-powered technology can now capture information, coordinate logistics, process inventory, prepare taxes, provide financial services, translate complex documents, create business reports, draft legal briefs, and diagnose illnesses. Moreover, thanks to machine learning (ML), these tasks are expected to improve significantly in the coming years. Powered by big data, computers can learn, practice skills, and ultimately improve their own performance and perform assigned tasks more efficiently.
Our new research paper uses data from 24 European countries to assess the ‘automation potential’ of various jobs. This probability is first calculated at the job level and then aggregated at the occupation level (Table 1). Since each job is made up of different tasks with different automation possibilities, the potential for automation at the job level does not necessarily mean the destruction of jobs, but rather whether automation can significantly change the nature of those jobs.
Table 1: European jobs with the highest and lowest potential for automation
Source: Brekelmans and Petropoulos (2020) based on Nedelkoska and Quintini (2018).
We use this means of automation in an aggregation framework where jobs are divided into three different skill categories: low-skill, medium-skill, and high-skill. Figure 1 shows the results.
Figure 1: Exposure to automation of different skill groups
Source: Brekelmans and Petropoulos (2020).
These results suggest that artificial intelligence and machine learning will have a different impact than computer and robotic technologies, which have caused job polarization (a decline in medium-skilled, routine jobs and an increase in low-skilled jobs). In contrast, AI is likely to significantly transform not only medium-skill jobs but also low-skill employment. Additionally, while highly skilled talent is at relatively low risk of being transformed by AI and ML, the impact on these occupations remains significant.
The results also suggest changes to the future of work. For medium and low-skill jobs, AI systems complete tasks that can be easily automated, while humans continue to perform tasks that cannot. Increased productivity through the introduction of AI technology may also create new tasks and jobs with a high probability of automation, but these jobs and tasks are likely to require advanced skills.
The transformative nature of AI and ML requires proactive measures to redesign the labor market. Countries with more flexible workforces, high-quality scientific education, and less pervasive product market regulation tend to adopt more skill-based job structures and are therefore less exposed to workforce transformation through automation.
The transformative nature of AI and ML requires proactive measures to redesign the labor market. Employees need to prepare for upcoming changes while also taking advantage of the efficiencies that these technologies can provide. Countries with more flexible workforces, high-quality scientific education, and less pervasive product market regulation tend to adopt more skill-based job structures and are therefore less exposed to workforce transformation through automation.
This blog was created within the “Future of Work and Inclusive Growth in Europe” project, with funding from the Mastercard Center for Inclusive Growth.
Recommended citation
Brekelmans S., G. Petropoulos (2020), “The major impact of artificial intelligence on low- and medium-skill jobs.”Bruegel’s blogJune 29, available at https://bruegel.org/2020/06/artificial-intelligences-great-impact-on-low-and-middle-skilled-jobs/
