This paper presents simultaneous models that predict induced travel demand and induced road investment using an array of instrument variables reflecting political, environmental, and demographic influences. From a panel data set consisting of 22 years of observations for 34 California urban counties, short-run elasticities are estimated. Both the Vehicle-Miles-Travelled model and the Lane-Miles model feature good statistical fits and highly significant parameter estimates. While the research found strong reciprocal relationships between road investment and travel demand, the elasticity estimates generated from simultaneous equation modelling were generally of a comparable magnitude to those of past single-equation studies. (Author/publisher).
Abstract