Employment and Social Developments in Europe (ESDE) 2023

2. Labour shortages, including skills shortages, in the EU

There is no universally agreed methodology on how to measure labour shortages in the EU. One of the most common approaches is to use the BCS, which collects quarterly survey data from employers in the industry, manufacturing and services sectors, asking whether labour is a major factor limiting their production. ⁽¹⁰⁷⁾ Often, job vacancy rates are used as an alternative measure to identify demand for labour, as they cover more sectors, albeit with less sectoral disaggregation and with a bigger time lag. The European Labour Authority (ELA) also collects information from the public employment services (PES) on labour shortages and surpluses, by occupation and Member State. However, that information is more qualitative than quantitative, and the time series is short. This section combines all three approaches to map the sectors and occupations with persistent labour shortages in the EU – referred to as ‘labour shortage sectors’ and ‘labour shortage occupations’, respectively. The available approaches, however, do not provide a single reliable estimate of the size of labour shortages, i.e. the number of vacancies that cannot be filled due to labour shortages.

Skills shortages are often approximated by the share of employers reporting difficulties in finding employees with the right skills. Firms’ recruitment difficulties may be driven by various factors. These can include the lack of skills among job applicants, ⁽¹⁰⁸⁾ but also poor recruitment, retention ⁽¹⁰⁹⁾ and HRM practices, ⁽¹¹⁰⁾ high replacement demand due to retirement, the need to fill new jobs created as a result of the green and digital transitions, or companies unwilling or unable to offer competitive wages and working conditions (see section 7.). Recruitment also varies over the business cycle, as employers tend to be more selective (e.g. raising skills requirements in their vacancies) when the pool of unemployed people is larger. ^{(111) (112)} Exploratory research based on employer interviews suggests that employers are often unable to define and measure the specific skills they are looking for when recruiting. ⁽¹¹³⁾

Analysis of European Company Survey (ECS) 2019 data ⁽¹¹⁴⁾ shows that the lack of skills among job applicants is not the main reason for companies’ overall difficulties in finding people with the right skills. Rather, the likelihood of experiencing recruitment difficulties is strongly driven by the inability to retain workers, ⁽¹¹⁵⁾ with its impact being 10 times larger than that of skills requirements (at 14.2 pp and 3.4 pp, respectively) (Table 2.1). ^{(116) (117)} There is also an institutional dimension to firms’ recruitment difficulties, with the presence of a recognised body of employee representation associated with a reduced chance of experiencing difficulties in recruiting workers. Recruitment difficulties are less likely in larger firms, which are better able to recruit from a larger available talent pool (both externally and internally), as well as more likely to have an employee representation body. ⁽¹¹⁸⁾ To shed more light on the skills demanded in occupations facing persistent labour shortages in the EU, this section analyses skills needs based on the European Centre for the Development of Vocational Training (Cedefop) European skills and jobs survey (ESJS2) 2021 data. ⁽¹¹⁹⁾

2.1. Identifying persistent labour shortages and their skills needs

The first approach uses BCS data and identifies persistent labour shortages in construction, manufacturing, information and communication technology (ICT), transportation and storage, professional, scientific and technical activities, and administrative and support services. In order to identify subsectors with persistent labour shortages by country over time, several methodological assumptions ⁽¹²⁰⁾ were made about the length of time series, the threshold at which labour shortages were considered severe, the threshold for considering severe labour shortages persistent, and the criteria for identifying subsectors with persistent labour shortages at EU level. ⁽¹²¹⁾ Table 2.2 presents the full list of subsectors (NACE ⁽¹²²⁾ 2-digit level and the corresponding NACE 1 level) identified as facing persistent labour shortages across the EU over the last 10 years (2012-2021). ⁽¹²³⁾

The second approach uses Eurostat job vacancy rates and identifies persistent labour shortages in human health and social work activities. This sheds light on the labour demand in some sectors that are additional to the ones identified by the BCS such as human health and social work. Since 2012, overall vacancy rates ⁽¹²⁴⁾ are high and increasing in arts and entertainment, administrative and support services, construction, information and communication, professional, scientific and technical activities, accommodation and food services, and transport and storage (Chart 2.1). The vacancy data support the findings of the BCS and also show increasing shortages in human health and social work. A sharp decrease in overall rates was observed in 2020 due to the COVID-19 outbreak, with the exception of healthcare. From 2021, the rates increased and reached historical highs for all sectors except mining and quarrying. Looking at the latest quarterly data at the end of 2022, the overall vacancy rates slightly decreased but remained historically high, at 2.8% (Chapter 1). A significant drawback of this measure is that it does not distinguish whether high job vacancy rates in a sector are driven by high turnover or by labour shortages.

Total employment in the subsectors experiencing persistent labour shortages has increased by 11% since 2012, according to the EU-LFS (Table 2.2). The share of vacant posts in the sectors is increasing (Chart 2.1), indicating greater labour demand. Total employment is also high, due to the expansion of these economic sectors. The highest employment increases are recorded in computer programming (+79%), architectural and engineering activities (+24%), and social work activities (+30%) (Table 2.3). In 2021, approximately one-third of employed people, including the self-employed, were working in shortage subsectors. Increasing employment, coupled with reported shortages and vacancies in the specific group of sectors, provides evidence for persistent shortages over time.

The third approach lays the foundation for the analysis of structural drivers of persistent labour shortages at occupational level. It identifies persistent labour shortages for health, ICT professionals, personal services, sales, personal care, building and related trades, metal, machinery and related trades, and electrical and electronic trades workers, as well as drivers and mobile plant operators, and cleaners and helpers (Table 2.3). The approach is based on the occupation classifications developed by the International Labour Organization (ILO) and data collected by EURES national coordinating offices in the Member States. Since 2016, these offices have provided annual data on shortage occupations for European-level reports on labour shortages and surpluses. ^{(125) (126)} In order to identify labour shortage occupations at International Standard Classification of Occupations (ISCO) 3-digit level ⁽¹²⁷⁾ in the EU over time, certain assumptions were made and two criteria created. ⁽¹²⁸⁾ However, shortage occupations at 3-digit level might include some occupations at 4-digit level without persistent labour shortages. At the same time, the selection criteria mean that some occupations with persistent labour shortages at 4-digit level might be excluded from a shortage occupation group when aggregated to 3-digit level. ⁽¹²⁹⁾

The required skills differ across occupations with and without persistent labour shortages. Based on ESJS2 2021 data, labour shortage occupations have a higher frequency of workers requiring manual skills and lower values of literacy, numeracy and digital skills than non-shortage occupations (Chart 2.2). ^{(130) (131)} Descriptive analysis shows that the differences between labour shortage and non-labour shortage occupations are less pronounced on social skills, routinisation, and job complexity.

Even when controlling for different characteristics of occupations facing persistent labour shortages, labour shortage occupations are still more likely to be comprised of jobs with lower skill demand (Table A.3). ⁽¹³²⁾ This might explain the relatively low impact of skills in driving firms’ recruitment difficulties. Labour shortage occupations are more likely to experience skills gaps. Once other factors are taken into account, employees in labour shortage occupations perform routine tasks less frequently and are more likely to exercise greater discretion in organising and planning their work and in learning or adapting to unexpected situations or varying tasks. This suggests that persistent occupational shortages might be an outcome of firms’ human resource practices and their expectations of finding workers with high non-cognitive skills. ⁽¹³³⁾ The findings also suggest that there is scope to further develop the skills of those employees to move towards better job performance.

There are significant differences across broad occupation skill groups in the skills needed (Table A.3). ⁽¹³⁴⁾ For example, skilled shortage occupations typically rely more on physical rather than cognitive tasks, compared to skilled non-shortage occupations. However, they tend to be characterised by higher upskilling and digital skill needs, higher demand for better-educated workers, and greater job complexity, compared to skilled non-shortage occupations. Semi-skilled and manual shortage occupations are more likely to depend on less routine and more social tasks, compared to the same non-shortage skill groups, with a higher demand for lower-educated workers. In addition, semi-skilled shortage occupations depend more heavily on manual skills, while elementary shortage occupations primarily involve highly routinised tasks. All groups, with the exception of skilled shortage occupations, require a relatively basic or very low digital skill level, implying that a lack of digital skills is not the main driver of many persistent labour shortages.