The multidimensional roots of social (dis)advantage: How structural and individual characteristics shape inequalities from school to adulthood

2026-06-04T13:59:24+03:00

Session room:

Mapping Intersectionality in Adult Skills Across Nordic-Baltic Countries: An Intersectional MAIHDA Approach
Ronny Scherer, University of Oslo, CEMO & CREATE, Norway

Although inequalities in adult foundational skills—numeracy, literacy, and adaptive problem solving—are well-documented, research has predominantly examined social group differences one axis at a time. This obscures how gender, age, migration background, educational attainment, and employment status operate simultaneously to produce compound patterns of (dis)advantage. Intersectionality theory offers a more realistic account of how social positions interact; yet quantitative applications in adult skills research remain scarce. The Nordic-Baltic countries present a particularly instructive setting: despite strong institutional commitments to equity, meaningful skill gaps persist, and whether these reflect intersectional concentrations of disadvantage is largely unknown. In this study, we drew on PIAAC Cycle 2 data and took an Intersectional Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (I-MAIHDA) approach to quantify intersectionality across the five social characteristics above. The Variance Partition Coefficient (VPC) quantified the degree to which skill variation was attributable to stratum membership. Country-specific estimates were synthesized meta-analytically to quantify between-country heterogeneity and identify possible explanatory variables. Four country-level moderators were examined: the EIGE Gender Equality Index, adult education participation rates, gender-by-migration employment gaps, and Inglehart-Welzel self-expression values. We further examined whether intersectionality patterns were correlated across domains, probing whether compound disadvantage operates as a general phenomenon or manifests differently by skill. We expect intersectionality to vary across both countries and outcomes, with adaptive problem solving showing the strongest stratification given its unequal distributional preconditions. Countries with greater gender equality and broader adult education access are hypothesized to show attenuated intersectional concentration. Cross-domain correlations are expected to be moderate, suggesting a shared structural basis alongside meaningful domain-specificity. The findings will address whether aggregate equity indicators adequately capture the distributional complexity of adult skills, and whether intersectionality in foundational competencies constitutes a coherent cross-domain construct. By integrating intersectional quantitative methods with cross-national meta-analysis, this study advances both the measurement and explanation of skills inequality—with direct implications for targeted policy investment. School Belonging in a Tracked Education System: Evidence from Czech PISA 2018 and 2022.

School Belonging in a Tracked Education System: Evidence from Czech PISA 2018 and 2022
Kaitlin Griffith, Charles University, Czech Republic

Using PISA 2018 and 2022 data from The Czech Republic, this study examines school belonging a highly tracked education system. The Czech Republic is a useful case because PISA students are distributed across several school types at age fifteen: basic schools, vocational schools, four-year gymnázia, and selective multi-year gymnázia. These tracks differ not only in academic orientation, but also in socioeconomic composition and students’ duration of membership in a given school setting. The study addresses three questions: (1) To what extent do school track, school socioeconomic composition, gender, and individual SES explain school belonging among 15-year-olds in the Czech Republic? (2) How is school belonging associated with academic achievement? and (3) Do these patterns remain stable across the 2018 and 2022 PISA cycles? The analytic sample included 6,884 students in 328 schools in 2018 and 8,209 students in 418 schools in 2022, after excluding special schools and cases missing key covariates or weights. School belonging was modelled as a latent construct using the six PISA belonging items. Two-level structural equation models were estimated in `lavaan`, with students nested within schools and student weights applied. Student ESCS was decomposed into within-school and between-school components to distinguish individual SES from school socioeconomic composition. Achievement models used all ten plausible values across reading, mathematics, and science, with estimates pooled using Rubin’s rules. Results show that school belonging is patterned more consistently by socioeconomic position and gender than by school track. Higher individual ESCS was associated with stronger belonging in both 2018 and 2022. School-average ESCS also predicted belonging in both cycles, indicating that students in more socioeconomically advantaged schools reported stronger belonging, net of their own SES. Gender was significant in both years: boys reported stronger belonging than girls after controls, and this adjusted gender difference was substantially larger in 2022. School-track differences were less stable. In 2018, students in multi-year gymnázia reported lower belonging than students in basic schools, while vocational and four-year gymnázium students did not differ significantly from basic-school students. In 2022, the pattern shifted: vocational and four-year gymnázium students reported stronger belonging than students in basic schools, while multi-year gymnázium students no longer differed significantly. These results do not support a simple peer-continuity explanation in which longer membership in the same school setting leads to stronger belonging.

Like Parent, Like Child? Intergenerational Occupational Persistence in 31 Countries
Marco Paccagnella, OECD, France

Using data from the OECD Survey of Adult Skills (PIAAC) we document the prevalence of intergenerational occupational persistence in 31 countries. We define occupational persistence as working in the same occupation as at least one parent when the respondent was 14, measured at the ISCO 1, 2 and 3-digit levels. Occupational persistence is widespread: 28% of adults are occupational followers, meaning that they work in the same broad occupation (ISCO 1-digit) of at least one of their parents. Narrowing the definition of occupations at the ISCO 2-digits, the share of occupational followers falls to 11%, and to 7% when occupations at the ISCO 3-digit level are considered. In all countries observed occupational persistence is higher than what would be observed under independence of occupational choices across generations. Occupational persistence is higher in asset- and family-based production and less frequent in fast-growing and emerging occupations such as ICT. Occupational inheritance is gendered and increases with parental education. On average, differences in earnings and job satisfaction between occupational followers and non-followers are small and not statistically significant.

Parallel session F

The multidimensional roots of social (dis)advantage: How structural and individual characteristics shape inequalities from school to adulthood