Heterogeneous Effects of Digital Infrastructure on Sustainable Economic Growth: Panel GMM Evidence from Developing Countries (2014–2025)

Abstract

Abstract

This study investigates the varying implications of digital infrastructure on sustainable economic growth in five major emerging countries (Egypt, India, Kenya, Saudi Arabia and Sudan) during the period 2014-2025. We use panel data, System GMM estimation and Random Forest regression and classification to study the impact of internet penetration, mobile broadband subscriptions and fixed broadband subscriptions on sustainable GDP per capita growth. Our results indicate the statistical and economic impact of digital infrastructure on economic growth. The biggest benefits are from internet penetration, then mobile broadband. Fixed broadband is not statistically significant in the full model, which reflects the low access in these countries. Random forest regression (R2=0.538, RMSE=2.31) and classification (F1-score=0.714, AUC-ROC=0.792) models can accurately predict 75% of high and low growth regimes in out-of-sample validation. Internet penetration drives growth in middle income countries (Egypt, India) while mobile broadband drives growth in low income countries (Sudan, Kenya). Average education >6 years doubles the benefits of digital infrastructure growth on human capital. The sustainability implications of these results are relevant for the SDG 4 (quality education), SDG 9 (infrastructure and innovation), and SDG 10 (reduced inequality) programs. There is a digital divide as the MENA countries have 87% internet penetration compared to 44% in Sub-Saharan Africa. We recommend the expansion of mobile broadband in low-income countries and investments in internet infrastructure and human capital in middle-income countries to maximize sustainable growth.