Factors influencing inter-modal facility location decisions: Comparison of different empirical estimation procedures

This study examines several different empirical models of inter-modal terminal characteristics using a combination of available data from the Surface Transportation Board waybill survey and Oak Ridge National Laboratories national freight transportation database in the estimation process. Empirical models included ordinary least squares, fixed-effects models, and random effects models. The data also presented issues of heteroskedasticity and autoregressive processes which are addressed in the paper. Due to the spatial context of the intermodal facilities, the data was also evaluated for spatial interactions and autocorrelation. Substantial evidence of underlying spatial relationships in the data were observed and noted. The results of this analysis identify those empirical variables and characteristics that contribute to the economic and operational sustainability of inter-modal facilities, and the relative importance of each factor. These assessments may then be used to evaluate infrastructure investment decisions, particularly by public entities. This methodology offers an unbiased framework for identifying and evaluating public and private benefits resulting from such investments, and serves as a guide for transportation policy involving inter-modal freight.