Project Untangled

Follow us on social media

Project Untangled




    Some skills are being made obsolete by automation, while at the same time the demand for people who can perform non-routine tasks is increasing, according to a policy brief from the three-year project, which examines the effects of globalisation, demographics and technological change on labour markets in Europe and beyond.


    Meanwhile, technology-driven changes are increasing demand for digital skills, and certain social skills, such as independence and enthusiasm, have also gained importance. Demand may shift again if artificial intelligence contributes to the automation of more non-routine skills. Employees’ self-perception of being under-skilled has also increased, the researchers find.


    “Technological progress increases the value of new skills that younger workers tend to have an advantage in and diminishes the value of older skills that more senior workers possess,” said Piotr Lewandowski, head of the Institute for Structural Research, a member of the UNTANGLED research consortium. “As population ageing reduces the number of labour market entrants in Europe, investing in the skills of prime-aged and older workers becomes essential.”


    To address the skill mismatch, the European Union named 2023 the European Year of Skills, with the aim that by 2030, 60% of adults will participate in training each year, increasing the employment rate to 78%.


    Despite the need for upskilling and reskilling, the number of adults in training is low in Europe, ranging from slightly over 25% in Nordic countries to less than 10% in Eastern Europe (measured for the four weeks prior to the date they were surveyed).


    UNTANGLED researchers argue that when providing opportunities for adults to upgrade their skills and qualifications, companies and policymakers should take into account that employees’ training needs vary and depend on many factors including age, gender, family, current skills, industry, and where they live. Unfortunately, the available training is scarce and doesn’t always match these needs.


    “Data show that more than half of adults don’t see the point of training, suggesting that offerings should be improved, and more has to be done to raise awareness and outreach,” Lewandowski said. “Meanwhile there is also a group of employees who want to train but are stymied by time constraints and high costs.”


    UNTANGLED researchers advocate looking into successful programmes and initiatives, including the introduction of on-the-job training grants, Individual Learning Accounts, and Training Leave.


    Individual Learning Accounts, introduced in France, are like virtual savings accounts where balances accumulate over time that can be used to fund training, allowing workers to choose the skills that are most important to them. Over 2 million workers in France have used ILAs.


    Training Leave programmes operating in Sweden let workers take time off from their jobs to improve their skills. This allows overcoming the problem of not having enough time, which stops almost 30% of workers from training. Giving financial help to workers taking leave and offering longer leaves could encourage more participation.


    “Sweden’s long-term Training Leave programme addresses key obstacles employees face when thinking about training: it offers financial help and flexibility in choosing training,” said Wojciech Szymczak from the Institute for Structural Research, who co-authored the policy brief.


    Finally, Lewandowski and Szymczak suggest expanding on-the-job training opportunities in small and medium-sized companies. Smaller companies often can’t afford proper training for their workers; policymakers could help by covering the cost. For example, Dutch authorities successfully introduced a programme that reimbursed training costs for such companies, leading to over 180,000 training projects in 2019.


    Policymakers should take local conditions into account when implementing the recommendations, consulting with social partners such as trade unions and employers’ organisations, the authors say.


    Lewandowski, P., & Szymczak, W. (2023). UNTANGLED Policy brief: Fostering lifelong learning as the market for skills evolves (Deliverable 7.3). Leuven: UNTANGLED project 1001004776 – H2020.


    You can read the policy brief here.




    Arntz, who is also Leibniz Professor of Labour Economics at Heidelberg University, will discuss “Different Perspectives on the Digital Transformation of the Labour Market.”


    Registration for the conference, titled “Labour Market Effects and Social Impact of Technological Transformation, Globalisation, and Demographic Change”, is now open.

    Register here.

    Fore more information please visit our Events section.

    The full programme will be available soon.




    The research explores the extent to which wage disparities among employees in similar roles are influenced by a company’s willingness to invest in improving skills. The authors analysed spending both on general training and on IT-specific courses.


    The paper finds that employees at European firms that invest in training earn 9% more, on average, than their peers at companies that do not prioritise workforce development. Among IT-intensive workers, this wage gap expands to 17%.


    “Training is seen as a tool for improving employees’ opportunities and working conditions, and for increasing company productivity,” Cecilia Jona-Lasinio said. “The need for training increases with the pace of technological change. With the adoption of new technologies and acceleration of automation, formal education and experience are not sufficient, so new competencies are required.”


    While previous studies showed 16% of European workers are exposed to skill-displacing technological change, the paper from Jona-Lasinio and Venturini shows that the adverse effects of digitalisation can be tackled at the company level through training policies.


    The authors analysed data from 112,000 companies, employing between 10 and 999 workers, from twelve European countries, and found that 65% provided general training, with the highest proportions observed in France (84%) and the lowest in Bulgaria (26%). One-third of the companies also offered training in advanced digital skills, with the highest number in the UK, Norway, Germany and Denmark.


    Data show that firms which invest in training tend to be bigger. The highest wages are paid by organisations where spending on upskilling is the highest, and where the largest share of the workforce is trained. In terms of the type of training, i.e. internal or external offered by a specialised institution, the combination of both yields the best results, as it gives employees a mixture of company-specific skills and more general ones.


    “On an individual level our study shows that people who want to earn more should be ready for lifelong learning, and search for employers who offer training,” says Francesco Venturini. “Our study also suggests that the wage differences across firms might widen if laggard companies were unable to systematically organise training. We should keep this in mind when designing policies on innovation and support for companies.”


    Jona-Lasinio, C. and Venturini, F. (2023), “On-the-job training, wages and digitalisation: evidence from European firms”, International Journal of Manpower, Vol. ahead-of-print No. ahead-of-print.

    You can also access the paper here.





    The rapid expansion of automation has raised concerns for workers’ welfare, as robots are often implemented to improve efficiency and hence may replace humans. In the United States, the displacement of jobs due to automation has substantially reduced wages and employment, especially for those at the lower end of the wage scale.


    But in a study of 14 European countries between 2006 and 2018, a period of intensive robot adoption in the European Union, researchers found that tax and benefit policies largely absorbed labour market shocks caused by automation.


    “While the installation of industrial robots slightly decreased wages and reduced employment among the most exposed workers, the effect on household disposable income inequality was relatively small,” said Karina Doorley, a researcher at the Economic and Social Research Institute in Dublin, and a co-author of the paper. “Welfare and social benefits were a principal reason for this.”


    The study shows that while inequality has widened in many countries in the last 15-20 years, it was not driven by robot adoption, but by other factors and policy choices that limited income redistribution.


    Doorley and co-authors used data from several EU sources, including the EU Structure of Earnings Survey, the EU Labour Force Survey, and the EU Statistics on Income and Living Conditions to estimate the effects of robot penetration on wages and employment rates. Next, these estimates were used to calculate wages and employment rates that would have been recorded in 2018 if robot penetration remained at the 2006 level in each country. Finally, the resulting data was injected into EUROMOD, a tax-benefit microsimulation model, to assess the anticipated effects on household incomes.


    The automation-driven increase in household income inequality was most pronounced in Eastern European countries that witnessed increases in robot penetration, such as Slovakia and Hungary. However, the magnitude was small – in all countries studied, it was below 1.5% of the 2018 value of the Gini index, much lower than the actual changes in Gini indices in most countries between 2006-2018.


    Automation had a minor impact on household income inequality despite widening wage inequality and market income inequality to a larger extent. “In most countries, workers more exposed to robots tend to live with other household members with similar exposure. This pattern slightly amplified the automation shock at the household level, but the size of this effect was tiny, especially compared to the role of benefits,” the researchers wrote. They found that in general, benefit systems played a dominant role in cushioning market income losses, while taxes had a more muted role.


    Past studies into the drivers of income inequality have focused on the role of tax-benefit systems and demographic change, but this is the first evaluation to isolate the effect of automation, the authors said.


    There are some limitations to the findings. First, the simulations did not consider behavioural responses to automation. Second, changes in non-labour market incomes, fertility, or household structures were not included in the calculation.


    Still, the findings paint a less gloomy picture of robots’ role in Europe’s economic development, the author said.


    “Automation will affect the future of work, but robots explain only a minor share of changes in household income inequality in the countries that we studied,” said Piotr Lewandowski, a head of the Institute for Structural Research and co-author of the paper. “Tax-benefit systems were, and will continue to be, essential to mitigating the effects of automation shocks,” Lewandowski added.


    Doorley, K., Gromadzki, J., Lewandowski, P., Tuda, D., and Van Kerm, P. (2023). Automation and Income Inequality in Europe (Deliverable 3.5). Leuven: UNTANGLED project 1001004776–H2020.


    The paper is available here.




    Date: 19 October, 2023

    Time: 2:00-3:45 pm CEST


    This online event, titled “Harnessing the Benefits of AI, Industry 5.0, and Other Digital Innovations – More Opportunities but Also More Challenges for Social Innovations?” is organised by the European School of Social Innovation (ESSI) and will be chaired by Antonius Schröder (TU Dortmund University).


    Don’t miss out! Register by 13 October here 

    For more details and the programme, please visit ESSI website




    Incentive pay schemes (IPS) are remuneration policies that link an employee’s salary to their performance and the achievement of specific objectives. An increasing number of companies are adopting these policies to motivate employees and enhance organisational performance. While IPS tend to widen wage disparities within firms, scholars and policymakers have expressed concern that their growing popularity might exacerbate the gender pay gap. However, Perugini and Pompei’s study reveals that companies implementing IPS actually experience a narrower pay gap between female and male employees. Nevertheless, the effect is absent in companies that invest heavily in technology and intangible assets.


    The researchers analysed the salaries of over 6 million workers employed in 142,251 companies across Germany, France, Italy, Spain, and the UK. They found that female workers, who constituted 47% of the sample, earned on average 12.5% less than their male counterparts. The firm-level gender pay gap was most pronounced in the UK, at 16%, while it was narrowest in Italy, at 9.5%. When combining this data with information on IPS adoption, Perugini and Pompei discovered that a higher prevalence of IPS correlated with a narrower gender pay gap.


    “There are a number of factors that can trigger a wage gap, and in our analysis we controlled for their impact,” said Perugini. “In companies with performance-based remuneration the pay gap shrinks. This is probably because these companies attract a specific type of female worker, that is women who have similar skills, potential and ambitions to men, and who face fewer constraints related to household responsibilities. Additionally, these employers tend to have more effective screening and monitoring tools to prevent discrimination.”


    Furthermore, Perugini and Pompei found that IPS are more commonly adopted by innovative organisations, where incentivising worker performance is crucial for success. However, the positive impact of IPS on narrowing the gender pay gap varies by industry and is related to investment in information and communication technology (ICT) and intangible assets. The pay gap narrows primarily in firms that invest less in technology.


    The authors highlight that more substantial investments in ICT, research and development (R&D), and brand development often result in unpredictable working hours, making it harder for women to balance work and household responsibilities. Consequently, their chances of meeting their targets decrease, and female employees earn less than their male co-workers.


    “Our study suggests that it’s not IPS themselves that worsen gender pay inequality, but rather the specific technological context in which they are implemented,” said Pompei. “This underscores the need for policy measures aimed at redistributing unpaid work responsibilities between genders, as proposed by the Work-Life Balance Directive.”


    Cristiano Perugini & Fabrizio Pompei (2023) Pay incentives, intangibles, and gender wage inequality, Industry and Innovation, DOI: 10.1080/13662716.2023.2254264



    2021 © UNTANGLED. All rights reserved.
    This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101004776

    Follow us on social media