Supplementing Fixed-route Transit with Dynamic Shared Mobility Services: A Marginal Cost Comparison Approach

PI: Qing Shen (UW),, ORCID: 0000-0002-0968-7377

Co PIs: Catherine Gifford

AMOUNT & MATCH: $40,000 from PacTrans; $40,000 Match

PERFORMANCE PERIOD: 3/16/2021 – 3/15/2022

STATUS: Active

CATEGORIES: Transit, Mobility Services



FINAL PROJECT REPORT: will be available once completed

PROJECT DATA: will be available once completed

DESCRIPTION: Based on KCM’s policy experiment and rich program data, this research aims at introducing a rigorous approach that compares fixed-route services and shared mobility options based on their marginal costs. It will address the following questions:

1) What are the differences between the marginal cost and the average cost of providing fixed-route transit service? What factors can help explain such differences?

2) How do marginal costs vary between performance-equivalent traditional transit service expansion (e.g., increasing bus frequencies on existing routes, adding fleet size, increasing route density) and on-demand shared mobility services such as Via?

3) How can public transit agencies empirically estimate and compare the marginal costs of different options across different locations within a metropolitan region using rigorous data-based approaches?

Our analytical framework consists of four components. The first is theoretical reasoning based on the economic concept of marginal cost. The marginal cost of running a fixed-route transit in low-demand areas can be substantially higher than the average system-wide or route-wide cost. This perspective may help transit agencies identify areas—especially low-demand areas—where shared mobility modes have real competitive advantages. The second component is analyzing data from the Via to Transit program by quantifying performance metrics (e.g., rider total travel time, rider waiting time, rider monetary cost). These metrics will serve as benchmarks for the third component, the agent-based modeling, which simulates scenarios when Via to transit were not available. There will be two types of agents in the model: riders and the transit agency that operates fixed-route buses. The agent-based model will reproduce interactions of the riders and the transit agency on a typical commuting day. Key outputs of the model will be the level of transit services and the number of parking spaces needed to maintain an equivalent performance for users as that provided by Via. The last component will use the outcomes of the agent-based modeling, combined with cost information from KCM, to determine the marginal cost of providing performance-equivalent transit service expansion to meet the demand of Via riders at each Link stations. We will also examine how the marginal costs vary across different stations, and what transportation, geographical, and socioeconomic factors can help explain such differences.

Research Project Progress Report #1 10/10/2021
No Cost Extension Request 1/15/2022
Draft Report 1/15/2022
Final Project Report 3/15/2022