We are doing science for policy
The Joint Research Centre (JRC) is the European Commission's science and knowledge service which employs scientists to carry out research in order to provide independent scientific advice and support to EU policy.
The digital transformation of the economy and society has intensified the need for digitally skilled labour force. Recent studies inform about expected increased demand, and skill shortages in the Information and Communication Technologies (ICT) sector, with a widening gap between supply and demand of ICT specialists. The need for accurate data on the number of ICT specialists in employment becomes more pertinent due to the development of policy initiatives aimed at increasing digital skills.
Eurostat and the OECD define ICT specialists and propose a statistical definition using the International Standard Classification of Occupations (ISCO) 2008. Based on the Labour Force Survey, Eurostat provides an estimated 8.9 million persons working as ICT specialists in 2018 in the EU. This indicator annually feeds the Digital Economy and Society Indicator, a composite indicator that assesses the digital performance of EU Member States.
This paper shows that this value underestimates the actual number of ICT specialists and proposes a more accurate method for the estimation. The list of ICT occupations includes both 3-digit (3d) and 4-digit (4d) codes. The number of EU Member States not reporting 4d data was 12 in 2011 and 6 in 2018. Therefore, the direct implementation of the definition is not possible, and a method is needed to estimate the missing 4d values and produce complete estimates for the EU. Eurostat developed an estimation method based on education data (EF method) to provide estimates for ICT in employment. This paper proposes the Ratio method for the estimation of missing data, compares its performance with the EF method, and produces estimates on ICT specialists in employment for 2004-2018, for the EU and its Member States.
The results show that the Ratio method provides more accurate estimates than the EF method. We test the performance with two error measures by means of a cross-validation algorithm; in both cases, all six variants of the Ratio method tested reduce the error of the EF method between 35% and 55% when measured on countries reporting 4d data. The new proposed method estimates 9.2 million ICT specialists are working in the EU in 2018, 2% above the value with the old estimation method (the difference reached 26% in 2004). At country level, for countries with missing data, the new method implies an average increase of 34% in 2004-2010 and 17% in 2011-2018 with respect to the estimate with the current method. According to our estimations, the number of ICT specialists in employment followed an increasing trend over the two analysed periods (2004-2010 and 2011-2018), with an overall increase of 19% in the first period, and of 35% in the second one. The share over total employment also increases, with a grow from 3.2% in 2011 to 4.0% in 2018. These results are in line with other studies that show that the ICT sector was more resilient to the economic crisis that started in 2008 than the whole economy.