Education, Cognitive Skill, Workplace Skill Use, and Employment Outcomes, Comparing Welfare Regimes Difference with a Multi-Group SEM Approach

Introduction

Education and skills are widely regarded as the main drivers of employment outcomes. Human capital theory assumes that individuals with higher skills obtain better jobs and higher earnings. However, growing evidence on skill mismatch and job polarization suggests that inequality depends not only on how much skill individuals have but also on whether their skills are utilized at work. Meanwhile, the skill-employment relationship operates not through universal mechanisms but through context-dependent pathways shaped by institutional configurations. This study examines how education, skills, and workplace skills use shape labor‑market outcomes, whether skills use mediates these relationships, and further examines whether these relationships differ across welfare regimes.

Method

This study uses data from the second cycle of the Programme for the International Assessment of Adult Competencies (PIAAC), a computer‑based survey of adults aged 16–65 conducted across in 2022–2023 (N=50,992). We examine how individual skills translate into labor‑market outcomes through workplace skill use and whether these pathways vary across welfare regimes. Monthly earnings and employment quality serve as outcomes. Key predictors include latent constructs of individual skill (literacy, numeracy, problem‑solving) and workplace skill use (technology, reading, writing, numeracy), with controls for age and gender. Structural equation modeling (SEM) is used to estimate mediation effects, and multi-group analysis is conducted to examine cross‑regime (Social Democratic, Liberal, Corporatist, East Asia) differences.

Result

The multi-group SEM demonstrated acceptable fit across four welfare regime groups (CFI=.950, SRMR=.043). The chi-square difference test confirmed that structural path coefficients differed significantly across welfare regimes (Δχ² =1548.8, Δdf=39, p<.001), supporting the institutional moderation hypothesis. Across all regimes, workplace skill use was the most consistent and powerful predictor of employment quality (β=.654–.739), effectively mediating the impact of cognitive skills. Notably, cognitive skills alone had minimal direct effects on earnings (β=.058–.112), indicating that "having" a skill is insufficient without the opportunity to "apply" it. Direct returns to education (credentials) were highest in Liberal regimes (β=.256), and workplace skill use had the highest impact on earnings in Corporatist regimes (β=.409).

Discussion

This analysis challenges the traditional human capital assumption that simply enhancing individuals’ skills will yield better employment outcomes. Instead, the ability to use skills in the workplace emerges as a more decisive factor, highlighting the importance of job–worker matching and the organizational conditions that enable employees to apply their competencies. While active labor market policies often prioritize upskilling and reskilling, these findings suggest that improving opportunities for skill utilization within existing positions may be consequential for labor‑market outcomes. Regarding welfare regimes, they shape how education and skills are converted into labor‑market advantages. Direct credential returns are strongest in Liberal regimes and East Asia, consistent with market‑driven signaling in which educational certificates function as key productivity cues. Corporatist regimes exhibit the highest returns to workplace skill use, potentially reflecting stronger institutional support and collective bargaining structures that reward applied competence. Taken together, the results suggest that enabling workers to apply their skills and understanding regime-specific institutional contexts are both crucial for explaining cross-national variation in labor-market returns.