From autonomy to digital leadership: Multilevel pathways to school improvement in OECD education systems

This study examines the association between principals’ decision autonomy, digital leadership, and perceived school improvement across OECD education systems. Using data from the TALIS 2024) principal questionnaire (N = 22,450; 29 countries), we estimate within-country associations using regression models with country fixed effects and design-adjusted standard errors. Digital leadership is examined in relation to decision autonomy, digital competence, technological barriers, and institutional stability. We conduct moderation analyses (Autonomy × Competence; Autonomy × Barriers), mediation analyses, cross-level interaction models, and structural equation modelling. Results indicate that decision autonomy shows the strongest positive association with digital leadership after controlling for country fixed effects (β ≈ .34), alongside positive relationships with digital competence (β ≈ .28) and negative relationships with technological barriers (β ≈ −.18). Digital leadership is positively associated with perceived school improvement (β ≈ .27), and mediation analysis indicates a statistically significant indirect relationship between autonomy and perceived improvement through digital leadership (indirect β ≈ .09). The findings indicate that decentralised decision authority is associated with more favourable organisational conditions when accompanied by principals’ digital competence and supportive technological infrastructures.