Do Smart Cities Grow Faster?

Previous studies have found a strong positive correlation between human capital, measured as the share of the adult population with a college degree, and population growth in metropolitan statistical areas (MSA) in the U.S. In this paper, I corroborate that the human capital-growth connection is indeed statistically significant, although much weaker than previously thought. The evidence suggests that the main reason behind this bias lies on endogeneity issues that have not been thoroughly addressed in the literature. In particular, omitting lagged MSA growth in regressions of current MSA growth on human capital overestimates the impact of skills by 100 per cent. Given that past growth has been shown to be one of the main drivers of current MSA growth (Glaeser 1994a), omitting the former variable in growth-education regressions would bias our human capital estimates upwards. Upon further examination, however, I show that MSA-specific fixed effects explain away the alleged impact of past on current growth. This suggests that the individual characteristics of the city that made it grow in the first place, and not lagged MSA growth per se, are what drives future MSA growth. Yet, even after accounting for these MSA-specific fixed effects, the impact of human capital on MSA growth does not disappear: my estimates suggest that a decadal increase of 10 per cent in the share of the adult population with a college degree translates into a rise of between 3 and up to 5 per cent in the MSA population growth rate during the same period. Finally, instrumental variable regressions strongly support the direction from skills to growth, abating potential reverse causality concerns.Clasificación JEL: R11, J24.