Digital Versus Traditional Capital Expenditure and Digital Revenue in Telecommunications: Panel Evidence From 13 Operators and the “Infrastructure Paradox” in Uzbekistan

This study examines whether digital capital expenditure contributes more to the digital revenue of telecommunications operators than traditional network investment, and explains why incumbent operators in emerging markets have expanded their networks rapidly while the share of revenue generated by digital services remains low-a pattern termed the infrastructure paradox. The research uses a balanced panel of 13 operators over 2015–2025 (143 observations). Fixed-effects models with Driscoll–Kraay standard errors are estimated, and robustness is verified with a two-step System GMM estimator. Digital investment has a coefficient of 0.321 (p < 0.001) on the digital revenue share, about 4.7 times the effect of traditional capital expenditure (0.068); competition intensity (0.156) and regulatory quality (0.118) are also significant. System GMM estimates are consistent (0.309; Hansen J p = 0.412; AR(2) p = 0.287). A Paradox Index relating infrastructure growth to revenue growth is proposed; Uzbektelecom JSC records the highest value in the benchmark group (0.76), compared with 0.18 for Singtel. The findings suggest that reallocating capital toward digital services and strengthening market competition yield higher returns than further network expansion, which has implications for state-owned operators in Central Asia.