Digital Diffusion, R&D Intensity, and Adult Learning Participation: Panel Evidence from EU-27 Economies

While digitalization is expected to increase the demand for continuing workforce training in European economies, large-scale, cross-country panel data on macro-level relationships between participation in adult learning and digitalization remain insufficient. This study examines how digital diffusion, employment in the information and communication technologies sector, research and development intensity, and institutional quality are associated with adult learning participation across EU-27 member states. An unbalanced panel dataset covering the 27 EU member states for the period 2015–2023 was created from Eurostat and Sustainable Development Goals monitoring indicators. The empirical strategy includes cross-sectional dependence diagnostics, unit root and cointegration tests, fixed-effects estimation with Driscoll–Kraay standard errors, moderation analysis, and robustness checks based on alternative covariance specifications and first differences. The level-fixed-effects results show positive associations between individual internet use, as a proxy for economy-wide digital diffusion, and research and development intensity with adult learning participation (β = 0.110, p < 0.01; β = 2.451, p < 0.01, respectively). Employment in the information and communication technologies sector is not statistically significant, and institutional quality does not significantly moderate the association between individual internet use and adult learning participation. In hypothesis terms, H1 (digital diffusion) and H3 (research and development intensity) are supported in the level specification, whereas H2 (information and communication technologies employment) and H4 (institutional quality moderation) are not supported. When standard errors are clustered at the country level, the key coefficients lose significance, so these level estimates are descriptive structural associations rather than causal effects. The first-difference estimator does not reproduce the level relationships, indicating that the findings primarily reflect cross-country structural associations rather than within-country dynamic effects. Country-group analysis reveals a fourfold difference in mean adult learning participation between the lowest and highest digital diffusion quartiles (5.3% vs. 20.3%). The findings inform EU digital and skills policy by suggesting that the expansion of digital infrastructure should be coordinated with adult learning targets within the 2030 Digital Compass framework.