Policy-makers and researchers are increasingly relying on online job advertisements (OJA) – also called “job postings” – as a new source of data to understand the dynamics of labour markets, the changing demand for skills and jobs. This type of data can provide an up-to-date picture of the skills that employers require, at higher detail than conventional surveys. The collection of reliable online job advertisement data is part of “Skills intelligence”, one of the actions of the European Skills Agenda.
In scientific research, OJA data is also increasingly used to understand the changing composition of tasks and skills of jobs over time, and infer the impact of automation and artificial intelligence on employment. In principle, this data could also be used monitor how the twin transition and other major trends are reshaping labour markets in the EU, by creating new occupations, making others vulnerable and changing the nature of existing jobs. However, the origin and processing of this data also presents statistical and analytical challenges.
The JRC analyses job advertisement data, assesses its reliability by comparing them to established data sources on tasks and skills, such as digital skills, and studies how employers advertise positions and express their skill needs.
Because OJA data is a new data source, we need to understand how representative and reliable they are: not all new jobs are advertised (online, or at all), so online data may provide a distorted picture of the labour market. Job advertisements are distinct from “job vacancies”, which denote unfilled positions that may or may not be publicly advertised, which are measured through surveys. By contrast, OJA are continuously collected from the internet, and encoded into structured dataset by encoding in terms of standard classifications, such as those for occupations, educational qualifications, and skills.
The JRC cooperates closely with Cedefop and Eurostat, the European Institutions collecting online job advertisements for the European Union through Skills-OVATE, as well as Lightcast, the leading global commercial provider for this type of data. JRC research also informs the development of ESCO, the European classification of occupations, skills, and qualifications, managed by the European Commission, used as the a reference taxonomy for the statistical analysis of online job advertisements.
Jobs, skills, and tasks
JRC research shows how online job ads data is skewed white-collar professional occupations, which are relatively better represented in their numbers and in their variety of skills and tasks, relative to less qualified occupations.
Comparing the occupation task profiles from UK OJA with data with the JRC-Eurofound Task Database, allows to map the rich, but relatively unstructured information on skills used OJA in terms of tasks, measured in occupational surveys. This mapping from skills to tasks, supported by a theoretical framework, is achieved through a custom Skill-Task Dictionary. The approach allows to track task profiles of occupations as described both by job advertisements and from representative surveys, and to compare the two sources.
In terms of task content, online job advertisements emphasise intellectual and social tasks, relative to manual ones. Teamwork is more visible compared to less desirable aspects of work organisation, such as routineness and standardisation, or forms of control that reduce autonomy and latitude. Digital tools (which require digital skills) are also particularly prominent.
Digital skills in online job advertisements
Online job advertisements are particularly detailed in terms of digital skills, by requiring specific programming languages and software requirements, especially in digitally-intensive occupations in Information and Communication Technology (ICT). Such information can help in correctly measuring digital skills, contributing to a priority of the European Skills Agenda.
Research by the JRC uses OJA data from the United Kingdom to measure the demand for different levels of digital skills, in terms comparable to the European Skills and Jobs Survey (ESJS) by Cedefop. Overall, online job advertisement understate the prevalence of basic digital skills like computer literacy and office software compared to the ESJS, as employers do not mention them explicitly in job advertisements, perhaps because consider them self-evident. However, OJA data is more representative about the prevalence of advanced digital skills, including specialised software and the emerging field of artificial intelligence. While jobs that use advanced digital skills command a wage premium, and their skillset changes more often than most other jobs, they account for a relatively small share of overall employment. Future JRC research will measure the levels of digital across EU countries using the ESCO skills taxonomy.
Do employers know what skills they need?
Online Job advertisements are intended to signal the skills required by employers. However, employers themselves may not always be able to accurately describe the skills they require from their recruits.
An exploratory study commissioned by the JRC interviewed private-sector firms, to better understand how employers develop job descriptions, and how they define, communicate and evaluate their skill requirements.
The findings suggested that firm size and the content of work are crucial to the process. Smaller firms in particular – which constitute most of EU employment – may not have formal job descriptions or be able to articulate their skill demands. The study also found that employers require “soft” and “transversal” many job profiles, but may be unable to define or assess them, and that these requirements may carry potentially problematic messages about worker personality or behaviour.
Future JRC research will explore how firms are using and measuring and assessing skills, in what context and for what purposes.
Publications
Rodrigues, M, Fernández-Macías, E., Sostero, M., A unified conceptual framework of tasks, skills and competences, JRC Working Papers on Labour, Education and Technology 2021/02, European Commission, Seville, 2022, JRC121897
Sostero, M. and Fernández-Macías, E., The Professional Lens: What Online Job Advertisements Can Say about Occupational Task Profiles, JRC Working Papers on Labour, Education and Technology 2021/13, European Commission, Seville, 2022, JRC125917.
Sostero, M., and Tolan, S., Digital skills for all? From computer literacy to AI skills in online job advertisements, JRC Working Papers on Labour, Education and Technology 2022/07, European Commission, Seville, 2022, JRC130291.
Goulart, K. S., Rodríguez-Menés, J., and Caroz Armayones, J. M., Job descriptions, from conception to recruitment: A qualitative review of hiring processes, JRC Working Papers on Labour, Education and Technology 2022/06, European Commission, Seville, 2022, JRC131040.
