Larysa Hrytsenko, Наталія Шелєміна, Dmytro Latysh
Digital transformation has become a critical driver of public sector efficiency and health system performance, yet its role in enhancing the effectiveness of health expenditure remains insufficiently understood. This study aims to evaluate how digital government development influences health outcomes and to determine whether digitalization improves or constrains the effectiveness of health expenditure across countries. The analysis is based on a panel dataset covering 191 countries over the period 2003–2022 (2,069 observations), using fixed-effects models with interaction and non-linear specifications, complemented by Driscoll–Kraay standard errors to ensure robust inference. The results reveal that digitalization exerts a strong, statistically significant direct effect on health outcomes, with a one-unit increase in the E-Government Development Index (EGDI) associated with approximately a 30–31% reduction in DALYs (β ≈ −0.305 to −0.310, p < 0.001). In contrast, health expenditure demonstrates a positive elasticity (β = 0.0628, p < 0.001), indicating that a 1% increase in spending is associated with a 0.063% increase in DALYs, suggesting inefficiencies or reactive spending patterns. The interaction between digitalization and health expenditure is positive and highly significant (β ≈ 0.465, p < 0.001), implying that digitalization does not automatically enhance, and may even weaken, the marginal effectiveness of additional health spending. Furthermore, non-linear analysis identifies a U-shaped relationship between digitalization and health outcomes (β_EGDI ≈ −0.259; β_EGDI² ≈ 0.610, p < 0.01), alongside an inverted U-shaped moderating effect (β_interaction ≈ 0.315; β_interaction² ≈ −0.392, p < 0.05), indicating threshold effects and diminishing returns. These findings highlight that while digitalization directly improves health outcomes, its impact on expenditure efficiency is conditional, non-linear, and dependent on the maturity and integration of digital systems.