In assessing user response to cost and service changes, demand elasticities are useful tools. Current compilations of demand elasticities, however, are not helpful to scheduled transportation operators. The reason is that the range of an elasticity is too wide, and there is no practical guideline for picking an appropriate value within the range. Furthermore, they are often compiled for systemwide andaverage-day operations, whereas most analyses need to be performed on a route-by-route basis during peak or off-peak periods. A methodology to address this problem is presented with the objective of providing demand elasticities that are practical for patronage analyses in an operating agency. Through statistical analyses of spatial and temporal data aggregation, the methodology explains the differences among elasticity tabulations, and in so doing, provides insights into the variations among elasticity values. The results of this research include (a) guidelines for selecting an appropriate value of elasticity among the broad range of values (instead of simply taking theaverage or midpoint of the range) and (b) a method for converting the most commonly available elasticities (which are usually in aggregate form) to a more useful form, such as route-specific and time-of-day elasticities. The results have been demonstrated in a case studyof the transit system in york, pennsylvania. This paper appears in transportation research record no. 1328, Travel demand forecasting: new methodologies and travel behavior research 1991
Samenvatting