Identifying literacy and numeracy skill mismatch in OECD countries using the job analysis method

Abstract

The Programme for the International Assessment of Adult Competencies (PIAAC) is currently the most important data source that provides information on the key skills possessed by workers, including literacy and numeracy. However, to assess skill mismatch, we also need information on the required skills in those domains, measured in the same metric and scale. In this article, we use the Job Analysis Method (JAM) to determine the required skill levels of literacy and numeracy for all four-digit International Standard Classification of Occupations (ISCO) 2008 unit groups. JAM is often considered the ‘gold standard’ in estimating required levels but has never been used in the context of PIAAC. This article thus presents the first results on the prevalence of skill mismatches using JAM comparing different OECD (Organisation for Economic Co-operation and Development) countries and different occupations. We compare our results with alternative methods and show some advantages of using JAM. We discuss some of the limitations as well.

Systematic Literature Review: An Analysis of Skill Mismatch Measurement

“…Advantages Disadvantages Realized Method Approach (Quintini, 2011) (Maltseva, 2019) The comparison of cognitive skills (literacy, numeracy, and problem-solving) with attainment value of occupation; Level of education required for the job (Flisi et al, 2017) Use International Standard Classification of Occupations; Measured with competency bandwidth (under-skilled and well match (Senkrua, 2021) More objective description of skills (OECD, 2013:5) Sensitive to a cohort effect, misleading education mismatch, less sensitive to outlier and technological change, allow only one education level to be appropriate for each occupation, and too broad occupation grouping and self-report data from PIACC (OECD, 2013) Uses only 1 digit of ISCO (to achieve enough good matches) (Pellizzari & Fichen, 2013) Job Requirement Approach / Direct Measurement (Maltseva, 2019;Senkrua, 2021) Divided into four categories of skills (Quintini, 2011) Measured by the standard deviation (Allen et al, 2013) Biased as the respondent tends to overstate the skills used at work (Perry, Wiederhold & Ackermann-Piek, 2014) Skill used is not a necessary proxy for skill requirement, average skills are considered well-matched (Van der Velden & Bijlsma, 2017) Job Analysis / Job Evaluation Method (Nedelkoska & Neffke, 2019) Analysis of education and skills reported by expertise (Nedelkoska & Neffke, 2019) No information about an individual job, only average skills, and education that has been grouped and become a fixed requirement for an occupation, overrated level of education compares to self-reported (Van der Velden and Van Smoorenburg, 1997) Time-consuming (Rodríguez et al, 2021) Expensive & not available at the national level and need recurring updates (McGuinness & Pouliakas, 2017) Amongst the listed measurements above, there is no agreement on the correct or exact way to measure skills mismatch (Nedelkoska & Neffke, 2019). As there are pros and cons to each measurement, the combination of each approach is the most recommended solution for measuring skills mismatch (Desjardins & Rubenson, 2011).…”

“…However, Nedelkoska andNeffke (2019), andRodríguez et al (2021) defined four approaches where three are the same as those mentioned by Quintini (2011), except for the direct measurement of skills. (Allen & Van der Velden, 2001;Green & McIntosh, 2007) Measure directly employee's skill in performing the job and training provided (Allen & Van der Velden, 2001) Measurement error in which employees tend to exaggerate or overestimate their abilities (Hartog, 2000) Employer Survey/ Linked Employer-Employee survey/Selfreported approach (Maltseva, 2019) Ease of data collection (Senkrua, 2021;Maltseva, 2019) Bias towards specific answers, small scale (specific industry and occupations) (Senkrua, 2021;Maltseva, 2019) Objective / Direct objective measurement…”