Infrastructure-Dependent Returns to Education: Evidence from Electricity Access and Economic Growth

This paper examines why identical investments in human capital generate divergent growth outcomes across countries. Using a global panel of up to 193 countries observed from 2000 to 2021, we investigate whether the growth returns to education depend on basic infrastructure conditions, proxied by electricity access. Employing a fixed-effects framework combined with likelihood-based threshold exploration, we find that the growth effect of schooling varies systematically with electrification levels. While the likelihood function suggests a reference threshold around 40 percent electricity access, the confidence set is wide, indicating that complementarities emerge gradually rather than at a sharp cutoff. Regime-based estimation reveals that schooling has no statistically significant impact on GDP growth in low-electrification environments, but becomes strongly growth-enhancing once electricity access is sufficiently high. A Wald test rejects equality of the schooling coefficients across regimes, providing direct evidence of infrastructure-dependent returns to education. These findings highlight that human capital accumulation alone is insufficient to drive economic growth and underscore the importance of coordinated investments in core infrastructure to unlock the productive potential of education.