-
Notifications
You must be signed in to change notification settings - Fork 100
Transit Pass Ownership
Work-in-progress
This plan is for a person level transit pass ownership model that will be developed in cooperation with SEMCOG. See Phase 6a Task 4 for more information on the task scope.
Transit pass ownership is defined as persons who purchase or are provided transit passes or subsidies.
The purpose of the transit pass ownership model is to represent more realistic fare in mode choice decisions, to represent the likelihood of different types of workers to receive a transit pass, and to test policies around increased transit pass ownership/participation.
Development of a generic transit pass ownership model is more difficult than first imagined because of several issues: there is heterogeneity in pass type, heterogeneity in discounts and subsidies, potential interaction effects with other mobility options, and effects of pass ownership on mode choice. In addition, most models developed to date are pass ownership models and not transit fare discount/subsidy models, which make the model more generic / transferrable.
The MAG transit pass model is an extension to the auto ownership model and is a function of geography, gender, presence of children, person type, age, income, auto ownership, and accessibility measures. The model assumes the pass discounts fare by 50%.
The DaySim transit pass model is a function of person type, income, transit station accessibility, transit accessibility to work location, and aggregate accessibility measures by person type. DaySim also has additional fare discount fractions available, which are in addition to pass ownership, and are useful for targeted policy analysis. These are hard coded and are by age bin.
The proposed model design is at the person level, before auto ownership since it avoids a joint model structure with auto ownership and allows for the effect of subsidized passes to affect auto ownership. The transit pass ownership model will include the following alternatives: fully subsidized, partially subsidized, yes but not subsidized, and no pass. The model will include household, person, accessibility to work/school, and pass cost discount variables.
More information on the model design is in Joel's proposed design presentation.
There is limited transit pass data in the 2015 SEMCOG household survey and the transit on-board survey "tour-based" person sample. There is limited transit pass data in the MWCOG household survey and the SANDAG household survey. Both surveys do not collect the amount of benefit / subsidy. The MWCOG transit on-board survey collects the type of subsidy but not the amount.
Thus, the availability of data to estimate model parameters varies by region. There is no good information on reasonable elasticities to the cost of transit passes and it is probably not possible estimate cost coefficient due to cross sectional data.
The limited data on fare discount/subsidy means Joel will do some further analysis / investigation of available data to inform a future discussion on the submodel design.
- Assume transit owning persons are insensitive to cost at the trip level?
- Some transit passes are only good for certain systems\operators, which could be problematic in regions with multiple operators.