Do person trip rates vary across contexts?

By Kristina M. Currans

In August 2014, the Institute of Transportation Engineers released the 3rd edition of the Trip Generation Handbook, a 352-page text that has traditionally, until only recently, provided guidance on estimating vehicle trips generated from new development. Among other updates, this new edition includes new chapters summarizing the most recent research and methods developed allowing users to account for people (not just vehicles) in trip generation estimation practices. This industry’s transition to estimating and understanding the “people” traveling to development has been in high demand from communities looking to accommodate multimodal travel, but there still remain a number of limitations in the guidelines presented.

One such limitation is that we currently have few data that allow us to directly estimate person trip rates. Instead we are often required to: (1) estimate vehicle trips using suburban vehicle trip rate data, defined as a “base rate”; (2) convert the vehicle trips into an estimated person trips using an assumed mode share and vehicle occupancy rate from the ITE’s suburban vehicle-oriented data, and; (3) reallocate the estimated person trips into different modes (bike, walk, drive, transit, etc.) based on the urban context of the development. This “direct mode share adjustment” forces the user, first, to assume that person trip rates do not vary across urban contexts and, second, to assume that person trip rates at very urban locations are the same as very suburban locations.

For a simple gut check, if we compare the person trips we would expect a restaurant in very urban downtown area to generate with a similarly sized restaurant located in a vehicle-oriented suburb, we would not expect these establishments to generate the same foot traffic through the door. In fact, we might hypothesize that the accessibility of a dense urban downtown would attract far more people to visit the establishment than in a suburban area, most of whom would be coming by walking, biking, or transit. Don’t these urban businesses pay more per square foot in part because the greater accessibility of these locations attracts more customers through the door than more exurban or suburban locations?

This “does not vary” assumption inherent in many of the “direct mode share adjustments” in the new methods and guidelines is problematic because it forces us to ignore the potentially high number of non-automobile trips in very urban areas supportive to multimodal travel, which were historically ignored by more traditional trip generation estimation methods. By ignoring these trips, we may also ignore the necessary provisions and mitigations we need to accommodate the multimodal traffic the urban locations generate.

Kristina Currans is a graduate research assistant at Portland State University's Oregon Modeling Collaborative. More information on the topic is in the Journal of Transport and Land Use, "Adjusting ITE's Trip Generation Handbook for Urban Context," which summarizes this TREC project.