Productivity growth, scale economies, and capacity utilization in US railroads, 1955-74.

Proposes a methodology for obtaining estimates of productivity growth that do not reflect changes in scale and movements toward or away from equilibrium. Attempts an unambiguous definition of productivity growth and demonstrates how its measurement can be approached from either the primal or the dual characterization of the production structure. Applies this methodology to obtain estimates of productivity growth for Class I US railroads in the period 1955-74. The estimates of the variable cost model show that returns to scale are sizeable for US railroads if output changes are accompanied by changes in haul and trip lengths. Concludes that the behavioural assumptions underlying cost function analysis have important implications for the measurement of productivity growth. -from Authors