Answers are grounded in this paper and cite their sources.
Overeducation is commonly measured using job analysis (JA), realized matches (RM), and self-assessment (SA). Although these approaches are frequently treated as alternative operationalizations of the same phenomenon, questions remain about the extent to which they identify the same workers and reflect a common underlying construct. Using data from the Czech Programme for the International Assessment of Adult Competencies (PIAAC 2022–2023; N = 2,742), this study systematically compares the prevalence, overlap, latent structure, and validity of alternative overeducation indicators. JA benchmarks were derived from the Czech National System of Occupations, while RM benchmarks were constructed from Czech Labour Force Survey data. Results reveal substantial variation in prevalence estimates across measures and only moderate agreement in individual classifications. The strongest convergence is observed between JA and RM, whereas subjective overeducation exhibits considerably weaker alignment with both objective indicators. Latent-structure analyses provide limited support for a single overeducation construct underlying all three measures. Instead, JA and RM form a highly overlapping pair of objective indicators, while SA captures a partially distinct dimension of educational mismatch. Validation analyses show that objective mismatch indicators significantly predict subjective overeducation but explain only a limited share of variation in workers’ subjective assessments. Profile analyses further demonstrate that workers identified by different combinations of indicators differ systematically in their characteristics. The findings challenge the common practice of treating alternative overeducation indicators as interchangeable and suggest that the principal distinction lies between objective and subjective approaches to measuring educational mismatch.
No takes yet. Share an insight, caveat, or question.
Petra Raudenská (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: