How Important are Human Capital, Physical Capital and Total Factor Productivity for Determining State Economic Growth in the United States, 1840-2000

This paper introduces a new data set of state-level physical capital in the United States from 1840 to 2000. The new data is combined with measures of the labor force, human capital, land, and output by state to perform traditional accounting decom- positions. Growth in measured inputs accounts for three-fourths of output growth, but variation in the level or growth of measured inputs accounts for only one-third of the variation in the level or growth of output. Our methodology is comparable to that used for cross-country accounting studies, and yields a quantitatively similar conclusion: most cross-state and cross-country variation in levels or growth rates is driven by variation in TFP. One interpretation is that while states have less institu- tional and technological heterogeneity (and consequently less TFP variation), this is matched equally by less input heterogeneity, with the relative importance of TFP and inputs remaining constant throughout the process of output per worker convergence. In this case, the relative importance of inputs and TFP may constitute a robust fact that applies across a variety of settings.