ABSTRACT
In 1980, Ann F. Friedlaender and Richard H. Spady (FS) published a pathbreaking paper on the U.S. freight market. Unlike earlier ad hoc studies of freight demand, their paper derived analytically the demand for truck and rail freight services from an underlying general cost function. In their analysis, shipper decisions were modeled using a translog cost function whose arguments included shipper outputs, their quasi-fixed stocks of capital and material inputs, labor prices, and generalized prices of truck and rail freight transportation. From the cost function, FS derived input demand functions and price elasticities for rail and truck transportation. Their study then drew on a cross section of industries in 1972 to estimate their model. This paper provides an updated analysis that draws on more recent data, and which provides new insights while using the same analytic foundation. The updated analysis presented here draws on a panel of observations from the 1997, 2002, and 2007 Economic Census reports and Commodity Flow Surveys. The estimated own- and cross-price elasticities for truck and rail are consistent with those reported by Friedlaender and Spady but significantly larger in absolute value. The results illustrate the extent to which U.S. freight railroads have been able to compete with trucking firms in markets that they did not serve effectively in 1972. Several explanatory hypotheses are offered.
