ICT Diffusion and Labor Productivity in G20 Countries: Development-Stage Differences and Granger Evidence

Abstract

This paper examines whether information and communication technology (ICT) diffusion is associated with labor productivity in G20 countries, whether the estimated relationship differs between advanced and emerging members, and whether ICT contains predictive information for subsequent productivity changes. A balanced panel of 494 country-year observations covering 19 G20 countries over 1998–2023 is constructed from the World Bank's World Development Indicators, with labor productivity measured as gross domestic product per person employed in constant 2021 purchasing-power-parity terms, ICT diffusion proxied by individuals using the internet as a percentage of population, and gross fixed capital formation, trade openness, and net foreign direct investment serving as controls. The empirical strategy is guided by panel diagnostics: the Pesaran cross-sectional dependence test, the Im–Pesaran–Shin unit root test, the Westerlund cointegration test, and the Hausman specification test inform the choice of estimator, which leads to a fixed-effects model with country-clustered standard errors, supplemented by the Dumitrescu–Hurlin panel Granger non-causality test and six robustness specifications. A one percentage-point increase in internet penetration is associated with approximately 0.36 percent higher labor productivity in the full sample (? = 0.0036, p = 0.015), precisely estimated for advanced economies (? = 0.0025, p = 0.011) and marginally significant for emerging economies (? = 0.0038, p = 0.081). Dumitrescu and Hurlin panel Granger non-causality tests on first-differenced series offer qualified asymptotic support for temporal precedence running from ICT diffusion to labor productivity. The pooled estimates conceal economically meaningful development-stage differences revealing implications for differentiated digital policy, institutional readiness, digital workforce capacity, and complementary investments in skills and technological capacity across country groups with mixed development.

Keywords: ICT diffusion, labor productivity, G20 countries, panel data, development-stage differences, Granger causality