Higher Education and Government Artificial Intelligence Readiness: A Cross-National Analysis
Governments' institutional capacity to harness artificial intelligence (AI) varies markedly across countries, and much of the existing literature attributes this disparity to digital infrastructure and income level. This study examines whether the accumulation of human capital through higher education explains additional variance in government AI readiness beyond these structural factors. Drawing on a cross-sectional panel of 158 countries that combines World Bank indicators, governance estimates, and a composite government AI readiness index, robust regression models (HC3), a non-parametric test of between-income-group differences, and a residual correlation analysis were estimated. Gross tertiary enrollment significantly predicted institutional AI readiness (β = 0.19, p = .001), even after controlling for GDP per capita, internet penetration, and government effectiveness. The marginal return of higher education on readiness was significantly smaller in low-income countries than in high-income countries, suggesting an absorptive-capacity threshold rather than a uniform linear effect. Countries whose educational performance exceeded what their income level would predict also showed, on average, higher-than-expected AI readiness performance (r = .32, p < .001). These findings qualify infrastructure-centered digital divide frameworks, engage with the literature on absorptive capacity and technological catch-up, and position higher education as a relevant, though not sufficient, institutional mechanism for explaining global inequality in AI governance.