Human capital and the rise of American cities, 1900–1990

Abstract We propose that cities that start out with proportionately more knowledgeable people grow faster in the long run because (a) knowledge spillovers are geographically limited to the city and (b) much knowledge is most productive in the city within which it is acquired. We found that city-aggregates and metropolitan areas with higher average levels of human capital grew faster over the 20th century. The estimated effects of human capital were large: a standard deviation increase in human capital in 1900 was associated with a 38% increase in average annual employment growth of city-aggregates over the period 1900–86. The estimated effects for metropolitan areas were smaller but still economically significant: a standard deviation increase in 1940 human capital was associated with an increase in average annual employment growth over the period 1940–90 of about 15%. Although the rise of the automobile appears to have overwhelmed the importance of human capital in cities dominated by manufacturing early on, human capital seems to have been economically more important in manufacturing cities than in non-manufacturing cities later on. Moreover, the estimated effects of human capital persisted for very long periods of time, suggesting either that adjusting to the steady state is very lengthy, or that shocks to growth are correlated with the presence of human capital.