This report aims to study the speed and direction of change that occupations undergo regarding skills composition for different European regions. One peculiar aspect of this research is the application of AI methods, such as word embeddings and machine learning to conduct economic analysis. We measure changes in the occupation skill sets based on the words used to advertise job vacancies in online job ads (OJAs) to measure changes between and within European countries, considering occupations’ specificity.
Which occupations will grow in the future and where? What skills will be demanded the most in the next years? How technology and digitalisation will affect existing and well-consolidated occupations? Those are the questions at the forefront of the policy debate among economists and policymakers. To address these questions, data-driven and real-time analysis of the labour market is needed to catch novelties - in terms of skills and new emerging jobs - as soon as they emerge from the labour market demand.
CESifo Working Paper No. 10288
We develop novel measures of early-career skills that are more detailed, comprehensive, and labor-market-relevant than existing skill proxies. We exploit that skill requirements of apprenticeships in Germany are codified in state-approved, nationally standardized apprenticeship plans. These plans provide more than 13,000 different skills and the exact duration of learning each skill. Following workers over their careers in administrative data, we find that cognitive, social, and digital skills acquired during apprenticeship are highly – yet differently – rewarded.
CESifo Working Paper No. 10281
We study how technology adoption and changes in global value chain (GVC) integration jointly affect labor shares and business function specialization in a sample of 14 manufacturing industries in 14 European countries in 1999–2011. Our main contribution is to highlight the indirect effect of robotization on relative demand for labor via GVC integration. To do this, we develop a methodology to separately account for robots in the total capital stock.
CESifo Working Paper No. 10237
This paper examines the labor market adjustments to four automation technologies (i.e. robots, communication technology, information technology, and software/database) in 227 regions across 22 European countries from 1995 to 2017. By constructing a measure of technology penetration, we estimate changes in regional employment and wages affected by automation technologies along with the reallocation of workers between sectors.
Papers in Evolutionary Economic Geography (PEEG) 23-02, Utrecht University, Department of Human Geography and Spatial Planning, Group Economic Geography
In this report we evaluate the opportunities for regional diversification in Europe over the last decade. We use microdata from the European Labour Force Survey to empirically test the entry and exit of occupational specializations at the regional level. Our results show that NUTS 2 regions are more likely to diversify into new occupations that are related to their existing local labour markets. So, the new opportunities for diversification are path-dependent, that is, they depend on the previous (occupational) production structure of the regions.
The paper examines the long-run versus short-run implications for labour markets of exposure to four automation technologies—robots, communication, information and software and databases. By applying a multiple break-point algorithm we identify investment cycles for each technology as affecting employment, wages, and wage shares for 163 NUTS-2 regions in 12 European countries over 1995-2017. In the long run, we find that robots have increased employment but reduced wages and the wage share in the region.
In recent decades, major secular trends in the labor market have significantly changed occupations and the skills demanded in occupations. In particular, advances in technologies and international outsourcing have decreased demand for certain types of tasks. Routine occupations are particularly vulnerable to automation risks, i.e., the risk that their tasks will be replaced by robots and automation technologies. Further, workers in occupations performing tasks that can be outsourced abroad face similar changes in skill demand due to lower trade barriers and technological advances.
There is a aggregate decline of manual-routine occupations due to substitution by automation capital, as these occupations perform tasks that can be easily replaced by machines. Similarly, technological progress and reduced trade barriers put occupations at an increased risk of offshoring, as their tasks can be performed abroad.
This report addresses is taking stock of the extant literature on the potential effects of technological change on labour outcomes. The paper focusses on the link between technological change, jobs and tasks. The paper makes a crucial contribution to existing reviews of the technology-employment nexus, by focusing on the technical and engineering literature, that describes the design of new technologies and how these execute tasks and jobs across industries.
Labor market developments such as globalization, structural transformation, and accelerating technological change can lead to mismatches between firms’ skill demand and employees’ skill supply. While skill mismatch is heavily discussed in research and policy, empirical evidence on the existence and determinants of skill mismatch in Europe is very scarce. In this paper, we develop novel measures of skill mismatch in Europe to address various questions of high relevance for labor market policies in the European Union: (1) How prevalent is skill mismatch in Europe?
CESifo Working Paper No. 10026
Immigration is one of the most divisive political issues in many countries today. Competing narratives, circulated via the media, are crucial in shaping how immigrants’ role in society is perceived. We propose a new method combining advanced natural language processing tools with dictionaries to identify sentences containing one or more of seven immigrant narrative themes and assign a sentiment to each of these. Our narrative dataset covers 107,428 newspaper articles from 70 German newspapers over the 2000 to 2019 period.