The conditional association between internet use and educational attainment, life expectancy and income per capita.

Abstract

Sub-Saharan Africa (SSA) has the lowest secondary school enrolment, life expectancy, and income per capita of any region, although internet use has grown rapidly. This dissertation examined whether the associations between internet use and these outcomes depend on electricity access and institutional quality. The study used panel data from the World Development Indicators and the Worldwide Governance Indicators for 2000–2023: 40 countries (956 observations) for education, 40 countries (920 observations, 2001–2023) for health, and 35 countries (840 observations) for income per capita. Fixed-effects models with Driscoll–Kraay standard errors and threshold interactions were estimated. For education, the conditioning hypothesis was supported, but not as expected. Internet use was positively associated with secondary school enrolment only when electricity access was high and institutional quality, proxied by government effectiveness, was below the median. For health, the conditioning hypothesis was not supported (p = 0.143). A positive static association between internet use and life expectancy did not survive a dynamic specification, the exclusion of Mauritius and Seychelles, or estimation for low-income countries alone. For income per capita, the conditioning hypothesis was supported. High institutional quality proxied by regulatory quality raised the elasticity of income per capita with respect to internet use in both models, in every income group and in all robustness checks. The interaction between internet use and electricity access was not robust. With year effects, the elasticities ranged from 0.013 (not significant) to 0.062. The associations differ by outcome and are sensitive to data coverage and specification. Connectivity programmes should be evaluated alongside regulation, electricity, and public services. Specifically, school-connectivity programmes should be piloted with electricity and evaluated before scaling up, while primary completion remains central to raising secondary school enrolment; digital health tools should be evaluated directly and routine immunisation protected; and network investment should be paired with regulatory reforms, such as easier licensing, enforcement of online contracts and competition among network operators. Keywords: Internet use, Institutional quality, Electricity access, Educational attainment, Life expectancy, Income per capita, Sub-Saharan Africa