The economic effects of the COVID-19 pandemic have been profound and widespread, prompting the examination of labour and skill needs and their alignment with training opportunities. Umbria, like many regions worldwide, is grappling with the challenges of digitalisation and shifting labour markets.

The COVID-19 pandemic severely impacted the Umbrian economy, and while labour demand has recovered, challenges like digitalisation, tight labour markets, and volatile demand for low-skilled jobs remain.

This chapter analyses the demand for labour in Umbria in between January 2018 and June of 2022 as measured by online job postings (OJPs). The analysis examines both long-term trends in demand, caused by factors such as digitalisation and sudden shifts due to the COVID-19 crisis in 2020. The chapter identifies high-demand and rapid-growth occupations and examines their characteristics using information contained in the requirements mentioned in OJPs. In particular, the analysis presents job characteristics such as educational and experiential requirements, as well as the types of contracts offered in job postings. Lastly, the chapter combines online job postings data with employment data obtained from the Labour Force Survey, to compare the prevalence of OJPs with that of employment figures and provide insight into labour shortages.

This chapter offers an overview of the courses provided in the Regional Training Catalogue (RTC) by the Umbrian regional agency for active labour policies (ARPAL). The analysis shows for which occupations training is available, and the corresponding number of training hours. Furthermore, leveraging Natural Language Processing (NLP) techniques, the chapter utilises algorithms and computational models to process and analyse the content of the courses described in the RTC in order to identify the skills that are provided in the training options available therein. Additionally, the chapter presents information on the cost, duration and class-sizes for the courses listed in the RTC, also highlighting the differences between the provinces of Perugia and Terni.

This chapter examines the alignment between the courses listed in the Regional Training Catalogue (RTC) and the demands reflected in online job postings, considering both sought-after occupations and skills. Utilising Natural Language Processing techniques, the chapter analyses the quantitative and qualitative match between the course content and the skill demand for each occupation included in the RTC. A novel metric, the skill-match score, is introduced by integrating data on sought-after skills from online job postings and the representation of these skills in the courses. Additionally, the chapter offers insights into potential areas where training may not yet adequately meet demand or exceed it within the analysed occupations and skill sets. These findings serve as preliminary indicators for policymakers, aiding in interventions to enhance training offerings or allocate resources accordingly.

This report uses a dataset of online job postings (OJPs) with monthly information between January of 2018 to June of 2022 to analyse Umbria’s labour market trends. The data is collected, transformed and harmonised by Lightcast (formerly Emsi-Burning Glass Technologies). The data is composed of 6.8 million individual level job postings for Italy and 72 434 for Umbria. There are up to 70 different variables ranging from skill keywords contained in each job posting, qualifications and experience required to fill the job and its geographical location, as well as the type of contract (permanent, temporary) and, when available, the salary offered for the specific role advertised. The OECD further transformed the data to create yearly aggregates, cross tabulations and other statistics presented in the document. Furthermore, the raw text of the OJPs is used for analysis, which is explained in this Annex. Lightcast offers the unique possibility to investigate the text contained in each online job posting, which reveals an amount of information that cannot be matched by any other source.