This paper presents new evidence on the social returns to education within a macroeconomic growth regression framework. I use improved schooling data and a macro version of the Mincer relationship between education and wages for individual workers. The results suggest that an increase by one year of the average education level of the labor force would increase labor productivity by 7-10% in the short run and by 11-15% in the long run. Some evidence is found for the presence of dynamic human capital spillovers: the human capital stock increases prospective economic growth. The empirical results are used to quantify the macroeconomic impact of skill upgrading as agreed upon in the European Union’s Lisbon strategy for growth and jobs. Finally, the paper discusses discrepancies between private and social returns to education.