This project deals with explaining the demand for public transport trips (or the sales of public transport tickets) by using aggregate time-series monthly data over a long-range time period (1973-1996). The research task is to test a two-step procedure, where the mode-of-payment (e.g. choice of ticket type) forms an aggregate logit sub-model, which is linked to the main demand model by a utility term. It also aims at testing various non-linear functional forms of the demand model. Recent research findings show a broad spectrum of explanatory factors affecting public transport patronage, not only prices and travel times. The project aims at analysing the development of transit ridership in the Stockholm and Gavle regions in Sweden. The aim is to estimate which variables (as endogenous variables we use price and supply, and as exogenous factors we use socio-economics and business cycle, climate and special events) influences the demand. By using non-linear statistical methods, the magnitude of these factors will be estimated. Similarities and differences in background variables and in the transit demand between the two regions will be examined. This paper deals with the Stockholm model, which is going to be estimated in the first quarter of 1998. The data base (more than 80 variables) and the first test model runs are already successfully carried out.
Abstract