Examining the role of civic attitudes in the link between family wealth and school dropout among tertiary vocational students

School dropout has serious individual and societal consequences, including poorer economic and health outcomes and loss of human capital. Dropout rates are particularly high in vocational education. Socioeconomic status (SES) reliably predicts dropout, with youth from lower SES families at greater risk. Adolescence is a developmental period in which youth become more aware of their family’s socioeconomic position and form civic attitudes (e.g., beliefs about fairness of society and trust in institutions) that can shape academic motivation and behavior. This study asks whether family wealth—captured as family affluence, perceived family wealth, and experienced financial scarcity—predicts school dropout among tertiary vocational students, whether civic attitudes (system justification and institutional trust) explain these links, and whether interpersonal trust in teachers can buffer against risks associated with lower civic attitudes. Understanding these mechanisms can reveal leverage points for interventions to reduce dropout and address social and educational inequalities.

Prior work shows a robust association between lower SES and higher dropout, with mechanisms spanning cognitive abilities, access to resources, behavior problems, and expectations. Research often emphasizes objective SES indicators (income, education, occupation) or affluence, but adolescents’ subjective perceptions (perceived family wealth) and experiences (financial scarcity) may be especially predictive of behavior and wellbeing. Adolescents from lower wealth contexts encounter more adverse environments and fewer supportive institutional resources, which may reduce system-justifying beliefs and institutional trust. Economic and social uncertainties strongly predict low institutional trust, particularly among younger people, potentially dampening engagement in success-related behaviors like school persistence. Positive teacher relationships can promote engagement and achievement and may buffer against low civic attitudes by modeling fairness and reliability; however, evidence on such buffering in vocational settings is limited. Mental health and migration background also relate to SES and dropout and warrant consideration as potential confounders.

Design: Preregistered (osf.io/ezjuf) prospective study linking baseline survey data (Sept 2019–Feb 2020) with official school dropout status about 9–12 months later (Oct/Nov 2020). Setting: Three tertiary vocational schools in the Netherlands (levels 2–4). Participants: 1,231 first-year students (Mage=17.81, SD=1.82; 44.31% male; 74.53% without migration background) from 71 classes; 81% participation of 1,519 eligible. Dropout data were missing for 172 students (13.97%) due mostly to lack of consent for administrative linkage. Measures: • School dropout: Officially reported status (1=dropout) in Nov 2020. • Family affluence: Family Affluence Scale (FAS-III; 6 items on material assets), summed and ridit-transformed to 0–1 (ordinal α=0.72). • Perceived family wealth: Single item “How well off is your family?” (1–5; reversed so higher=more well off). • Financial scarcity: Psychological Inventory of Financial Scarcity (PIFS; 6 items, 1–5; α=0.78). • System justification: 11-item system justification scale adapted to Dutch context (1–7; α=0.90). • Institutional trust: OECD 7-item scale (trust in politicians, police, healthcare professionals, government workers, news, courts/judges, social media information) rated 0–10; α=0.89. • Interpersonal trust in teachers: 3-item Teacher Support Scale (1–5; α=0.83). • Mental health problems: SDQ-R subscales for emotional symptoms (α=0.82), conduct problems (α=0.58), hyperactivity/inattention (α=0.79), peer problems (α=0.53); items scored 0–2, averaged and multiplied by 5 (0–10). • Background: Gender (0=female,1=male), age, migration background (0=both parents born in NL, 1=≥1 parent born abroad). Analysis: Path analysis in Mplus 8.8 with logistic regression for dropout. Stepwise modeling: (M1a–c) each wealth indicator predicting dropout; (M1d) all three wealth indicators simultaneously; (M2a–c) mediation via institutional trust (2a), system justification (2b), and both (2c); (M3) moderation by trust in teachers on civic attitudes-dropout links; (M4) add covariates correlated with dropout (gender, age, migration background, mental health subscales). Estimation via Maximum Likelihood with robust SEs (MLR) using full information, correcting SEs for class-level clustering (71 classes). Indirect effects via product of coefficients with 5,000 bootstrap replications for 95% CIs; total effects are sum of direct and indirect. Model comparisons used scaled chi-square difference tests and AIC/BIC. Missing data handled with full information MLR. Structural validity of multi-item measures acceptable; all models used manifest variables.

Sample and descriptive: n=1,231; 10.39% dropped out after first year. Financial scarcity mean below midpoint; perceived wealth, institutional trust, system justification around midpoint; trust in teachers above midpoint. Pairwise correlations: More dropout associated with lower family affluence, lower perceived wealth, higher financial scarcity, lower system justification, lower institutional trust, lower trust in teachers, and higher emotional, conduct, and hyperactivity/inattention problems (effect sizes small). System justification and institutional trust moderately correlated (r=0.53). Family wealth and dropout (Table 3): • M1a: Family affluence predicted lower odds of dropout (B=-0.890, p=0.026; OR=0.411). • M1b: Perceived family wealth not significant (B=-0.250, p=0.113; OR=0.779). • M1c: Financial scarcity predicted higher odds of dropout (B=0.358, p=0.015; OR=1.430). • M1d (all wealth indicators): Financial scarcity remained significant (B=0.303, p=0.025; OR=1.354) while family affluence became non-significant (B=-0.705, p=0.070; OR=0.494), perceived wealth non-significant (B=0.001, p=0.996; OR=1.001). Mediation by civic attitudes: Model fit improved when adding mediators, especially with both mediators (M2c). In M2c, the direct effect of financial scarcity on dropout disappeared, showing indirect-only mediation via institutional trust: higher financial scarcity → lower institutional trust → higher dropout (Bindirect=0.083, 95% CI [0.024, 0.156], OR=1.086). Total effect of financial scarcity remained positive (Btotal=0.289, 95% CI [0.034, 0.555], OR=1.335). Predicted probabilities illustrate total effect: dropout probability 12.69% at 90th vs 7.22% at 10th percentile of financial scarcity (>1.5x higher). Perceived family wealth showed an indirect-only effect via higher institutional trust → lower dropout (Bindirect=-0.055, 95% CI [-0.113, -0.012], OR=0.947) with near-zero total effect (Btotal=0.014, 95% CI [-0.218, 0.310], OR=1.014). Mediation via system justification alone (M2b) was observed for financial scarcity and perceived wealth, but these paths became non-significant when institutional trust was included (M2c). Moderation by trust in teachers: Adding trust in teachers as main effect and moderator did not improve fit; effects non-significant. Thus, links between civic attitudes and dropout did not depend on teacher trust. Confounding by mental health: Including emotional, conduct, hyperactivity/inattention, and peer problems (as relevant) improved model fit, but indirect effects via institutional trust remained comparable. Only hyperactivity/inattention significantly predicted dropout. Findings were robust to adjustment for mental health problems. Overall: Experienced financial scarcity is a stronger predictor of dropout than objective family affluence; institutional trust is a key mediator linking financial scarcity and perceived wealth to dropout; system justification did not uniquely mediate when accounting for institutional trust; teacher trust did not buffer effects.

Findings address the research questions by showing that adolescents’ experienced financial scarcity, rather than objective family affluence, is most proximally linked to dropout risk, and that this association operates through reduced institutional trust. This aligns with theories that economic strain undermines trust in institutions and reduces engagement in success-promoting behaviors such as school persistence. While system-justifying beliefs related to wealth indicators, they did not uniquely predict dropout once institutional trust was considered, suggesting a more proximal role for trust in specific institutions in shaping educational behaviors. The absence of buffering by trust in teachers suggests that, in this context and period (including COVID-19 disruptions), interpersonal teacher trust may not counteract low institutional trust or civic attitudes in predicting dropout, or that general teacher trust (vs trust in specific teachers) is less influential. The robustness of indirect effects after controlling for mental health underscores that mechanisms beyond emotional/behavioral symptoms—such as perceived control, short-term orientation, and beliefs about institutional support—may be critical. The results highlight institutional trust as a potential target for policy and practice to reduce dropout and mitigate socioeconomic inequalities.

This study advances understanding of SES-related dropout by distinguishing objective family affluence from adolescents’ perceptions and experiences of family wealth and by identifying institutional trust as a key mediator. Financial scarcity predicted dropout indirectly via lower institutional trust; perceived family wealth also showed a small indirect effect via institutional trust, while system justification did not uniquely mediate the association. Trust in teachers neither predicted dropout nor moderated civic attitude effects. Implications include the importance of strengthening adolescents’ institutional trust and addressing financial scarcity to promote school retention. Future research should: (1) test these pathways longitudinally across multiple waves to establish directionality; (2) examine generalizability across educational levels and countries; (3) refine measurement of family wealth (e.g., incorporate income, parental education/occupation) and multi-item perceived wealth; (4) differentiate types/directions of dropout (leaving education vs switching tracks); and (5) assess teacher-specific trust and ensure adequate power for interaction tests.

• Generalizability: Conducted in Dutch tertiary vocational schools; findings may not extend to other levels or countries. • Causality: Family wealth indicators and civic attitudes were measured at the same timepoint; associations are correlational and cannot confirm directionality. • Measurement of family wealth: Family affluence (FAS) is an adolescent-reported proxy and may differ from income; perceived family wealth was a single item. • Dropout outcome: Did not distinguish between complete exit from education and switching to other training pathways. • Teacher trust measure: Assessed general trust in teachers; trust in specific teachers may be more relevant; moderation analyses may have been underpowered. • Contextual factors: COVID-19 and related disruptions may have influenced financial scarcity and student–teacher relationships.