Blog Post

Artificial intelligence’s great impact on low and middle-skilled jobs

Artificial intelligence and machine learning will significantly transform low-skilled jobs that have not yet been negatively affected by past technological change.

By: and Date: June 29, 2020 Topic: Innovation & Competition Policy

The academic literature suggests that, in the past decades, technological progress has led to job polarisation in European Union countries. While computer technologies and robots have replaced, to some extent, routine middle-skilled jobs such as machine operation, construction work or administrative work, they have also led to an increase in complementary, non-routine high-skilled jobs (eg managers, professionals) and in low-skilled jobs (eg agriculture, cleaning and personal care services). However, our new research suggests that the new technologies that have emerged since 2010 – artificial intelligence and machine learning – are set to change drastically the job landscape over the next few decades. These technologies are likely to have a deeper impact across a wider range of jobs and tasks, including possible destruction of low-skilled jobs.

(…) new technologies that have emerged since 2010 – artificial intelligence and machine learning – are set to change drastically the job landscape over the next few decades. These technologies are likely to have a deeper impact across a wider range of jobs and tasks, including possible destruction of low-skilled jobs.

Artificial intelligence (AI) systems are able to perform tasks that involve decision-making, therefore changing the impact of automation on the workforce. AI-powered technologies can now retrieve information, coordinate logistics, handle inventories, prepare taxes, provide financial services, translate complex documents, write business reports, prepare legal briefs and diagnose diseases. Moreover, they are set to become much better at these tasks in the next few years thanks to machine learning (ML): computers fed by big data can learn, practice skills and ultimately improve their own performances and perform their assigned tasks more efficiently.

Our new working paper evaluates the ‘probability of automation’ for different jobs, using data from 24 European countries. This probability is initially computed at the job task level and then aggregated at the occupational level (Table 1). Since each job consists of a variety of tasks, with different potential for automation, the probability of automation at the job level does not necessarily mean the destruction of jobs, but rather whether automation can significantly transform the nature of those jobs.

Table 1: European jobs with the highest and lowest probabilities of automation


Source: Brekelmans and Petropoulos (2020) based on Nedelkoska and Quintini (2018).

We use this measure of automation in an aggregate framework where jobs are grouped into three different categories of skill: low, middle and high-skilled jobs. 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 different impacts compared to computer and robotic technologies, which caused job polarisation (drop in routine middle-skilled jobs and increase in low-skilled jobs). In contrast, AI is highly likely to significantly alter not only middle-skilled jobs, but also low-skill employment. Moreover, while the high skilled are relatively less at risk from AI and ML-induced transformation, its impact is still non-negligible for these jobs.

The results also suggest a future transformation of work. In middle and low-skilled jobs, AI systems will complete the easily automated tasks while humans continue to perform those that cannot be automated. A high probability of automation may also be associated with the creation of new tasks and jobs though the productivity gains from adopting AI technologies, but these jobs and tasks will most likely be high-skilled.

The transformative nature of AI and ML requires proactive measures to re-design labour markets. Countries with high degrees of labour flexibility, high quality science education and less pervasive product market regulations tend to have higher skill-oriented job structures and are therefore less exposed to labour transformation due to automation.

The transformative nature of AI and ML requires proactive measures to re-design labour markets. The workforce needs to be prepared for the upcoming changes, while the efficiency gains from these technologies should be harnessed. Countries with high degrees of labour flexibility, high quality science education and less pervasive product market regulations tend to have higher skill-oriented job structures and are therefore less exposed to labour transformation due to automation.

Recommended citation
Brekelmans S., G. Petropoulos (2020), ‘Artificial intelligence’s great impact on low and middle-skilled jobs’, Bruegel Blog, 29 June, available at https://www.bruegel.org/2020/06/artificial-intelligences-great-impact-on-low-and-middle-skilled-jobs/


Republishing and referencing

Bruegel considers itself a public good and takes no institutional standpoint. Anyone is free to republish and/or quote this post without prior consent. Please provide a full reference, clearly stating Bruegel and the relevant author as the source, and include a prominent hyperlink to the original post.

Read article
 

Blog Post

L'IA a besoin d'humains qualifiés

L'adoption des technologies IA repose moins sur des scientifiques de haut niveau que sur des spécialistes des données et des programmeurs compétents qui peuvent mettre en pratique les algorithmes d'apprentissage profond existants à des fins commerciales.

By: Julia Anderson and Guntram B. Wolff Topic: European Macroeconomics & Governance, Innovation & Competition Policy Date: October 5, 2020
Read article More by this author
 

Parliamentary Testimony

House of Lords

Employment and COVID-19

Testimony before the Economic Affairs Committee at the House of Lords, British Parliament on Employment and COVID-19.

By: Guntram B. Wolff Topic: European Macroeconomics & Governance, Finance & Financial Regulation, House of Lords, Testimonies Date: September 9, 2020
Read article
 

Blog Post

Europe has an artificial-intelligence skills shortage

How severe is Europe’s dearth of AI talent and how does it compare to the United States, China and the United Kingdom – the world’s AI champions?

By: Julia Anderson, Paco Viry and Guntram B. Wolff Topic: European Macroeconomics & Governance, Innovation & Competition Policy Date: August 27, 2020
Read about event More on this topic
 

Past Event

Past Event

The role of AI in healthcare

How can AI help us fight through a pandemic crisis?

Speakers: Dimitris Bertsimas, Georgios Petropoulos, Effy Vayena and Reinhilde Veugelers Topic: Innovation & Competition Policy Location: Bruegel, Rue de la Charité 33, 1210 Brussels Date: June 23, 2020
Read article Download PDF More on this topic
 

Working Paper

Occupational change, artificial intelligence and the geography of EU labour markets

From 2002 up to 2009, the economies of European Union countries went through a skill upgrading, rather than a polarisation between low-skill and high-skill jobs. After 2009, this changed, with declining real wages and a significant increase in the share of workers in low-skill jobs. This assessment evaluates these changes in connection with labour market variables, population densities and the emergence of machine learning and artificial intelligence.

By: Sybrand Brekelmans and Georgios Petropoulos Topic: Innovation & Competition Policy Date: June 15, 2020
Read article More on this topic More by this author
 

Opinion

Reading tea leaves from China’s two sessions: Large monetary and fiscal stimulus and still no growth guarantee

The announcement of a large stimulus without a growth target indicates that China’s recovery is far from complete.

By: Alicia García-Herrero Topic: Global Economics & Governance Date: May 25, 2020
Read article More on this topic More by this author
 

Blog Post

The European Union’s SURE plan to safeguard employment: a small step forward

The new EU instrument to mitigate unemployment risks during an emergency (SURE) is too modest to have a significant impact the COVID-19 crisis beyond being a first step in the overall recovery plan.

By: Grégory Claeys Topic: European Macroeconomics & Governance Date: May 20, 2020
Read about event More on this topic
 

Past Event

Past Event

Rationale and limitations of SURE

This event will discuss SURE, a new European Union instrument for temporary ‘Support to mitigate Unemployment Risks in an Emergency'

Speakers: Roel Beetsma, Elena Carletti, Grégory Claeys and Gilles Mourre Topic: European Macroeconomics & Governance Location: Bruegel, Rue de la Charité 33, 1210 Brussels Date: May 15, 2020
Read about event More on this topic
 

Past Event

Past Event

Find my virus: Mobilising AI and big data to fight COVID-19

At this event, the panellists will discuss the role of AI and big data in the fight against the coronavirus crisis.

Speakers: J. Scott Marcus, Alex Sandy Pentland, Georgios Petropoulos and Marietje Schaake Topic: Innovation & Competition Policy Location: Bruegel, Rue de la Charité 33, 1210 Brussels Date: April 2, 2020
Read article More on this topic More by this author
 

Blog Post

Artificial intelligence in the fight against COVID-19

Artificial intelligence can help fight the coronavirus through applications including population screening, notifications of when to seek medical help and tracking how infection spreads. The COVID-19 outbreak has triggered intense work on such applications, but it will take time before results become visible.

By: Georgios Petropoulos Topic: Innovation & Competition Policy Date: March 23, 2020
Read article More on this topic More by this author
 

Blog Post

Big data versus COVID-19: opportunities and privacy challenges

All available resources need to be brought to bear on the novel coronavirus COVID-19. To what extent can digital technology help? What risks are there in using big data to combat COVID-19, and what policies can mitigate any limitations that these risks impose?

By: J. Scott Marcus Topic: Innovation & Competition Policy Date: March 23, 2020
Read article More by this author
 

Opinion

Europe may be the world’s AI referee, but referees don’t win

The EU needs to invest in homegrown technology.

By: Guntram B. Wolff Topic: European Macroeconomics & Governance, Innovation & Competition Policy Date: February 19, 2020
Load more posts