This paper presents a model to estimate the rate of capacity utilization (CU) in the manufacturing sector by state. Consistent measures of state-level CU rates have been unavailable since 1982. The lack of a capacity measure has hindered regional studies of capital formation, long-run output growth, and productivity growth. Our model employs a neoclassical approach to estimate CU. We estimate a model to determine the optimal level of production and compare it with the actual level in order to define our CU index. Our results show that states in the West North Central, South Atlantic, and Pacific census divisions tend to have a CU index consistently above the nation's average. On the other hand, states in the East North Central and West South Central census divisions tend to have a CU index below the nation's average.

High labor costs in large Midwestern metropolitan areas have significantly reduced their manufacturing capital stock. For the period 1974 to 1978, the authors estimate that sixteen metropolitan areas in the Midwest, taken together, had approximately $2.8 billion less capital stock than they would have had if their labor costs had been at the national average. This difference is equal to 4 percent of the capital stock in these areas. The results are simulated from the estimation of a labor demand equation that is derived from a generalized Leontief cost function.

This paper seeks to reconcile the growth empirics technique of Mankiw, Romer, and Weil (1992) with the empirical results of Barro and Sala-“i-Martin (1991) through the development of a new database covering the 1977-96 period. We create state-by-state capital stock and gross investment estimates by apportioning the national capital stock among the states. Using these estimates along with gross state product and employment data, we find evidence that the Solow growth model explains state-wide growth during this period. We consistently find a rate of convergence of around 2%. Our results, as a consequence, suggest that the empirical results of Barro and Sala-í-Martin are driven by the neoclassical growth process of Solow.