Stochastic Frontier Models of Prism Vertices

Abstract

A methodology to estimate the location and size of space-time prisms that govern individuals’ activity and travel is presented. Because the vertices of a prism are unobservable, stochastic frontier models are formulated to locate prism vertices along the time axis using observable trip starting or ending times as the dependent variable and commute characteristics, personal and household attributes, and area characteristics as explanatory variables. Models are estimated successfully with coherent behavioral indications. A mean difference of 1.46 h is found between the observed trip ending time and the expected location of the terminal vertex for workers’ evening prisms. The estimation results aid in enhancing the understanding of prism constraints by identifying the determinants of prism vertex locations.