Digital transformation and entrepreneurship: A comparative analysis of EU11 and selected EU15 economies

  • 20 Views
  • 3 Downloads

Creative Commons License DMCA.com Protection Status
This work is licensed under a Creative Commons Attribution 4.0 International License

Type of the article: Research Article

Abstract
This study examines the association between multidimensional digital transformation and national startup ecosystem performance across 23 European Union economies, comparing EU11 with 12 selected EU15 economies from 2017 to 2024. The balanced panel contains 184 country-year observations. Principal component analysis is used to construct three composite indicators: the Digital Services Index, Digital Connectivity Index, and Digital Human Capital Index. The indices demonstrate satisfactory factorial adequacy, internal consistency, and one-component structures supported by parallel analysis. Their associations with startup ecosystem performance are estimated using two-way fixed-effects models. Because the panel contains only 23 country clusters, the principal inference uses CR2 bias-reduced country-clustered standard errors with Satterthwaite-adjusted degrees of freedom. In the full-sample specification, DSI has a positive but only marginally significant association with startup ecosystem performance (β = 0.0834, p = 0.082), while DCI, DHCI, and e-government are statistically insignificant. R&D intensity has a negative contemporaneous coefficient that is also marginally significant (β = −0.0145, p = 0.053). Regional heterogeneity is jointly significant and is concentrated primarily in digital connectivity: DCI is negatively associated with startup ecosystem performance in the EU11 group, whereas its total slope is approximately zero in the selected EU15 group. DHCI has a positive total slope within the selected EU15 group, although the difference between the EU15 and EU11 slopes is not statistically significant. A one-year-lagged specification produces a positive but marginal DSI coefficient and does not identify statistically significant associations for the remaining predictors. These estimates should be interpreted as conditional associations rather than causal effects.

view full abstract hide full abstract
    • Table 1. Definitions, measurement, transformations, and sources of the analytical variables
    • Table 2. Composition of the PCA-based digital indices
    • Table 3. Hypothesis-testing and empirical identification framework
    • Table 4. Descriptive statistics of the analytical variables
    • Table 5. Overall PCA and reliability diagnostics
    • Table 6. Baseline two-way fixed-effects model with CR2 inference
    • Table 7. Group-specific digital slopes and EU11–EU15 differences
    • Table 8. Joint CR2 test of the regional interaction terms
    • Table 9. Contemporaneous and one-year-lagged estimates with CR2 inference
    • Table 10. Summary of hypothesis testing
    • Table A1. Indicator-level PCA diagnostics
    • Table A2. Panel-model specification and diagnostic tests
    • Conceptualization
      Esmira Ahmadova, Lala Hamidova, Tetyana Nestorenko
    • Formal Analysis
      Esmira Ahmadova
    • Investigation
      Esmira Ahmadova, Lala Hamidova, Tetyana Nestorenko
    • Methodology
      Esmira Ahmadova
    • Project administration
      Esmira Ahmadova, Lala Hamidova
    • Resources
      Esmira Ahmadova, Lala Hamidova, Tetyana Nestorenko
    • Validation
      Esmira Ahmadova, Tetyana Nestorenko
    • Visualization
      Esmira Ahmadova, Lala Hamidova
    • Writing – original draft
      Esmira Ahmadova, Lala Hamidova, Tetyana Nestorenko
    • Writing – review & editing
      Esmira Ahmadova, Lala Hamidova, Tetyana Nestorenko
    • Data curation
      Lala Hamidova
    • Supervision
      Tetyana Nestorenko