Page 21 - IDEA Study 10 2015 Skills Mismatches

change of the regime will appear overeducated for their positions, because they will mostly work in occupations with majority of secondary-educated individuals. This mechanism might overestimate the share of empirically overeducated. Objective measures are based on formal dictionaries of jobs, compiled by job analysts who report different types of skill requirements for each occupation. These are then compared to the actual level of education of individuals who work in these occupations to assess the prevalence of overeducation. The objective measures of overeducation used in this study is the ET2020 Indicator of Mismatch used by the European Commission. This measure defines overeducation as share of individuals aged 25-34 with tertiary education that work in ISCO 4-9, i.e. not as legislators, senior officials, managers and professionals and not as technicians and associated professionals.11 The idea behind this measure is that tertiary educated individuals should work in one of the three highest ISCO categories (occupations with the highest skill requirements). If they work in occupations with lower skills requirements (ISCO 4-9), they are overeducated for the job they do. 5.2. Verticalqualificationmismatch:Results Main results for overeducation by subjective, empirical, and objective measures are reported in Table 2. Overall, about one fifth of the population finds themselves subjectively overeducated for their current job, while the empirical measure defines as overeducated only slightly more than 8% of the working population. Objective measure defines as overeducated about 12% of tertiary educated population. This is consistent with previous literature, which also finds larger extent of overeducation by subjective than by empirical and objective measures (Groot and van den Brink, 2000a). Overall, results of the three different measures of overeducation differ quite substantially, because each measure captures a different type of overeducation. Correlation between the three measures is thus relatively low (the highest correlation is between the empirical and objective measure – correlation coefficient of 0.7, but their correlation with the subjective measure is only 0.14 and 0.18, respectively). 11 For details, see https://crell.jrc.ec.europa.eu/ET2020Indicators/ 17