Modeling frequency of rural demand response transit trips

Abstract

Captive riders do not have many travel choices to meet the travel needs as fixed route transit services are not generally provided in rural areas. In many states, demand response transit (DRT) services are provided to meet such needs. However, state public agencies face the dilemma of whether to increase or decrease the service availability for on-call services. To enhance decision making of identifying what the causal factors related to DRT trips, the authors present a set of econometric models by integrating a sample DRT data with other explanatory variables such as land use, socio-economic, and demographic characteristics. Seven count data models including Poisson, Negative Binomial, Zero-inflated Poisson, Zero-inflated Negative Binomial (ZINB), Hurdle Poisson, Hurdle Negative Binomial, and ZINB Mixed Effect were developed to understand the factors that affect DRT trips. The ZINB Mixed Effect model that combines a zero-inflated negative binomial model with random effect was found to provide the best fit. A number of factors showed significant relationship with DRT trip frequency including distance, population density, elderly population, average income, and others. Further, the elasticity effects of these different factors were computed to quantify the magnitude of their impact on DRT. The proposed model can be helpful for transit agencies to predict the frequency of DRT trips and to provide adequate services in rural areas.