This article examines the long-run effect of higher education, measured in average years of tertiary schooling, on the level and growth rate of national per capita income. It uses an improved dataset on educational attainments which not only reduces measurement error but also overcomes data comparability issues and allows us to estimate the long-run effect of human capital through higher education on economic development. Using unique long panel data for 36 advanced and emerging economies spanning over the period 1870-2010, we find that higher education has a positive and significant effect on the level and the growth rate of national per capita income. Specifically, our empirical results indicate that a 1% increase in educational attainment at higher education level would raise the growth rate of per capita income by about 0.01% to 0.02% over a five-year period. That is, at the sample mean, an increase in average tertiary education by one year would raise output growth by about 6% to 11% in five years. We find that the returns to female education are lower than the returns to male education at tertiary schooling levels. Our results are robust to alternative specifications.