Activity-Based Models�& Economic Appraisal

UK Transport Modelers Forum

APRIL 22, 2026

Intro

• Economic appraisals are a common use case for travel models
• Activity-Based models (ABMs) are widely used in US regions
• Few applications of ABMs for economic appraisal
• ABMs present both benefits and challenges for economic appraisals
• How can ABMs become more widely used for economic appraisal?

Benefits of ABMs for Economic Appraisal

• Behavioral “resolution”
• Use of disaggregate synthetic population
• Heterogenous values of time
• Activity/purpose detail
• Greater spatial and temporal resolution
• Blocks, parcels
• Hours, Half-hours, Quarter-hours
• More detailed distribution impact analysiss

​

Challenges of ABMs for Economic Appraisal

• At what level should changes in utility be calculated (daily activities? tours? trips?)
• More models -> more parameters -> more sources of uncertainty
• Microsimulation / stochasticity / reproducibility / auditability
• Greater runtimes
• Complexity / communications challenges
• Development and maintenance costs
• Limited integration with existing economic appraisal methods

ABM System

5

AB Demand Model

Synthetic Population

Land Use

Transportation Network

Network Impedances (“skims”)

Supply / Network Model

Demographic Forecasts

Activity Patterns

HOME

WORK

HOME-BASED WORK TOUR

WORK-BASED

SUB-TOUR

HOME-BASED SHOP TOUR

SHOP

EAT OUT

INTERMEDIATE STOP

PERSONAL BUSINESS

ABM Flow & Outputs

7

• Lists of tours and trips for every resident, with detailed attributes
• Traffic volumes by roadway
• Transit boardings, alightings, and loads by stop and route segment
• Speeds and travel times

​

​

​

ABM Economic Appraisal Example #1: �Central Subway

• FTA New Starts program
• Federal program for transit guideway capital investments
• Multifactor evaluation framework
• Logsum-based measure
• Central Subway Project
• Extension of a new light rail line into a subway segment under downtown
• Goal: Improve service reliability, travel times and connectivity

​

​

​

​

FTA’s Transportation System�User Benefits Measure

• Quantify project benefits accruing to all transportation system across all travelers and travel modes
• Measures the change in consumer surplus
• Capture reductions in walk times, wait times, ride times, and number of transfers.
• Definition: the incremental annual cost of the project divided by all annual travel-related benefits in terms of hours saved by all users of the transit system
• Inputs
• Denominator of the logit mode choice model (tour and trip)
• In-vehicle time coefficient
• Incremental annualized capital costs plus annual O&M costs
• Measured by zone-pair for up to eight market segments

Source: Federal Transit Administration

Central Subway Results

Source: San Francisco County Transportation Authority

• Greater intensity of user benefits accruing at the trip level than at tour level
• Influence of rail ASC in trip model but not tour model
• Benefits reflect only service changes such as frequency and alignment
• Improvements in travel times offset by declines in accessibility due to fewer stop locations.
• Pros
• All travelers, all components of mode choice
• Can be produced by trip-based or activity-based models
• Cons
• Backwards / upwards linkages in ABM not captures by FTA imposition of fixed demand (in order to capture “first order” ie mode choice effects
• Holding the distribution fixed while varying mode leads to inconsistencies
• Need to calculate at the tour level and the trip level, without double counting

​

​

​

​

​

ABM Economic Appraisal Example #2: �MTC project Performance Assessment

• Utilizes the regional ABM logsum outputs
• Logsum captures mobility benefits (i.e., travel time savings, in-vehicle or out-of-vehicle, tolls, fares, parking, AOC)
• Converted to hours
• Monetized using a consistent value of time for all income classes (inconsistent with model assumptions, which incorportate distributed VOT)

MTC Benefit Cost Stream

Source: Metropolitan Transporation Commission

MTC Benefit Cost Results

ABM Economic Appraisal Example #3: �Ferry Planning

• Transit expansion planning
• Competitive funding application
• Benefit Cost Analysis
• Analysis approach
• Rule of Half

Source: San Francisco Bay Ferry

Mission Bay Ferry Project Description

• New ferry terminal enabling regional water transit expansion
• Serve a fast-growing urban area (San Francisco’s Mission Bay neighborhood)
• Project goals
• Alleviate (pre-pandemic) transit crowding
• Reduce area vehicle trips
• Earthquake transport resiliency

​

Source: calurbanist.com/san-francisco-bay-ferries/

Mission Bay Ferry BCA

• ABM-based Benefit Cost Analysis
• All behavior choice dimensions allowed to change
• Travel time savings
• Focused on trips between/to/from catchment areas of San Francisco ferry landings and regional destinations served by ferries
• “Rule of Half”: Simplified approach includes changes to in-vehicle travel time, walk time, wait time, and transfer time, but excludes changes to park-n-ride drive leg travel time
• Results
• NPV of $12M and BC ratio ~1.2
• Travel time savings address both typical weekday + special events
• Measurements sensitive to changing transportation services between opening and horizon year
• Auto time savings too small to differentiate from assignment convergence noise

ABM Economic Appraisal Example #4: �Transit Service Economic Impact Assessment

• Transit funding shortfalls due to post-pandemic ridership patterns and end of emergency funds
• Transit operators considering large service cuts
• Tax measures for transit operating funds on ballot
• Economic impact assessment of new taxes v. transit cuts
• Analysis approach
• Hedonic price model to estimate land value impact of transit service, based on ABM accessibilities
• Consumer surplus/deficit of travel accessibility changes due to service cuts (rule of half)
• Regional economic impact model to calculate citywide GDP implications

Benefits / Challenges / Next Steps

• Benefits / challenges
• What benefits are real? What are overstated?
• What challenges are real? What are overstated?
• Theoretically superior for capturing the behavioral responses
• Harder to integrate into appraisal frameworks designed around simpler behavioral assumptions
• Next steps
• What’s being done to address technical challenges?
• What are other barriers to wider adoption, and how can we address them?

​

Thank you

Joe Castiglione�Zephyr Foundation

joe.castiglione@zephyrtransport.org

​

Dan Tischler�San Francisco County Transportation Authority

dan.tischler@sfcta.org