Case Studies: Reviews Of Emerging Tools To Measure Equity –

Accessibility Observatory, Access to Opportunity Analysis

Interviewee: Andrew Owen, Director

Accessibility Observatory regularly evaluates access to opportunity on different modes of transportation in urban regions. Its popular “Access Across America” tool examines multi-modal access to jobs: how many jobs are reachable from Point A, using different modes of transportation and traveling at different times of day? The analysis uses Census jobs and demographic data, GTFS, and OpenStreetMaps, and its results are publicly available. Accessibility Observatory also conducts scenario planning for transit agencies, departments of transportation, and developers, evaluating gains in access to opportunity from potential projects.

Accessibility Observatory states that access to opportunity measures the supply of transportation and its intersection with land use, density, and housing – and thus the efficacy of the systems in getting people to and from destinations. Metrics can be disaggregated for different types of people, to measure how equitably a transportation system connects people to opportunities (though most analysis by Accessibility Observatory is not disaggregated). 

Accessibility Observatory notes that currently, access to opportunity metrics play a minute role in deciding if transportation projects are viable, and they are not used to evaluate if projects are equitable. This is partly because recent expansion of GTFS has made the metrics easier to calculate, and they are increasingly used in transportation decisions. Accessibility Observatory has expanded their services to train transportation officials on how to interpret access to opportunity data as it relates to project planning, and it hopes also to expand their data offerings to evaluate equity of access to opportunity.

Interviewee: Anson Stewart, Project Lead of Analysis and Research

Conveyal is a paid web-based software that computes access to opportunities  — an evaluation of effectiveness and equity in transportation networks —- on transit and other modes. It uses demographic and jobs data (with certain Census Bureau products pre-loaded), GTFS, and OpenStreetMaps. The platform simplifies the technical tasks to conduct the computations, thereby increasing the capacity for meaningful analysis by its transportation planner subscribers. 

Conveyal’s Analysis tool requires some technical skill, including familiarity with GIS, but analyzes access to opportunity with more sophistication and customizability than most other platforms. A few features make Conveyal Analysis unique: 

It can segment destinations by type of job and origins by residential demographics, producing results that imply disparities between groups, such as “The average low-income earner living in Boston can reach X low-wage jobs in 45 minutes on transit.”  The user can add additional parameters, including fare limits, number of transfers, or specified modes of transit. Users can also upload their own data sets. A map-based scenario editor tool enables transit scenario planning. The software is open-source, so anyone could replicate its methods. Conveyal updates its software regularly in response to client needs. The platform computes rapidly and includes a public engagement portal (Example: WeGo in Nashville) — both features enable planners to present impacts of service changes on access to destinations. This bolsters engagement, as riders can provide qualitative reactions to potential changes in real-time. Conveyal provides its subscribers with technical support, but it is not prescriptive about how results should be applied or interpreted. It can also contract with transit agencies to provide analytical support on scenario planning analysis. 

The Conveyal team recognizes deficiencies with the FTA’s Title VI Service Equity Analysis and therefore hasn’t built a TItle VI report generator. But one can essentially run a Title VI Equity Analysis by intentionally applying Conveyal features and thinking through results carefully. Conveyal looks hopefully to the growing popularity of access to opportunity, which it views as a better measure of transit equity than existing proximity measures.

Federal Transit Administration, STOPS Tool

Interviewees: Jim Ryan, FTA Planner; Jeff Roux, FTA Planner

STOPS is free software that models vehicle miles traveled and ridership. The FTA built STOPS to assist transit agencies applying for New Starts and Small Starts grants for fixed guideway projects — particularly small transit agencies without capacity to conduct the required build vs. no build analysis. 

The explicit transit focus, editable format, and ability to speed up lengthy forecasting analysis have made STOPS popular for broader transit modeling needs, and among transit agency staff, private planners, and researchers. STOPS can be reconfigured to represent changes in other transit modes (besides fixed guideway) and their impacts on ridership. The FTA is aware of STOPS being used to evaluate potential service changes, reimagine bus networks, do long-term planning, and measure equity of transit service. 

STOPS takes in commute pattern (from CTPP), GTFS, population, employment, and roadway travel times data; infers which trips will be taken given a change to the transit network; and outputs long-term ridership impacts. FTA staff recommend inputting rider survey data if available, which provides more certainty on rider demographics, behavior, and likely trips taken and ultimately creates more detailed and accurate forecasts. (Interviewees also noted that detailed, quality rider surveys are a good investment for transit agencies to improve their ridership modeling and equity analysis, in general.) 

STOPS in its base format separates ridership forecasts for zero car households, since the FTA evaluates if potential grantees increase ridership for this group. Editing STOPS code allows one to expand demographic inputs beyond car ownership, to allow more comprehensive analysis of the equity of potential changes to transit networks, but this step is technically cumbersome.

Florida Department of Transportation, TBEST Program

Interviewees: Rodney Bunner, Consultant/Developer; Chris Wiglesworth, FDOT Transit Planner; Matt Cheng, former Northern Virginia Transportation Commission Program Analyst

TBEST models short and mid-term changes to transit ridership, given proposed capital projects or service changes. The Florida Department of Transportation developed TBEST to increase the capacity of transit agencies, specifically to include accurate estimates in required planning documents. Its core usership are Florida transit agencies, but other agencies, MPOs, departments of transportation, academics, and private firms from across the country use TBEST. 

TBEST’s flexible nature means that it can automate many data analyses, enabling transit agencies to explore network changes and meet reporting requirements more easily. TBEST has a tool that takes data inputs and generates a report that meets the FTA’s Title VI Service Equity Analysis requirements. Users can modify the model to account for additional parameters, including how to define minority routes, service areas, disparate impact thresholds. TBEST users have also conducted scenario planning, explored how service tweaks impact ridership, and evaluated operating costs.

TBEST uses data from the Census, LEHD, MPOs, OpenStreetMaps, and GTFS feeds to establish existing conditions, to model how conditions will be impacted if the transportation network changes, and to produce report outputs summarizing those changes.TBEST is free but runs on the ArcGIS mapping software program. Users will need experience with ArcGIS and the software subscription to take advantage of the resource. TBEST is programmed with Florida data but it can be calibrated with inputs from other states; setup might take from 20 to 120 hours for non-Florida users. 

Alex Karner, Measuring Transit Equity 

Interviewee: Alex Karner, Assistant Professor of Community and Regional Planning at University of Texas at Austin

Alex Karner’s research explores transit equity metrics. According to Karner, singular focus on one equity indicator — as in the focus of the Title VI Service Equity Analysis on demographics of neighborhoods where service changes occur — flattens the nuance of the evaluation and leads to inaccurate conclusions about equity.

The Federal Transit Administration (FTA) sponsored Karner to create methods to evaluate equity in transit agencies’ systems. The resulting work (not published as of June 2021) expands industry practice to encompass new dimensions of methodologically-sound, replicable equity measurements. 

Karner developed four measures of transit equity that use widely-available data, vary in sophistication so that staff ranging in expertise can apply them, and capture the complexity of how transit service affects riders. To reach a robust conclusion about equity, an evaluation should apply multiple metrics and metrics should be disaggregated for people of different races, ethnicities, and income levels. 

Proximity to transit: who lives near a service change? The metric is similar to the Title VI approach and uses Census demographic data. Access to opportunities: how many opportunities (e.g. jobs) can someone reach on transit in an amount of time? The metric uses OpenTrip Planner and Census demographic and jobs data. Trip characteristics: what are travel times, transfers required, fares, or other characteristics of transit trips that people make? The metric uses rider survey data if available, or Census Transportation Planning Product (CTPP) data on commute patterns. Logsum measures: Given a transit service change, how do people’s travel times change? The metric is weighted by population and trips made. It uses FTA’s STOPS travel demand model and CTPP data. Karner’s FTA research also explored equitable approaches to planning and policy-making, including public engagement efforts, hiring practices, and community advisory committees.

Generally, Karner underlines that designing an equity analysis, interpreting its results, and “groundtruthing” its conclusions must heavily involve the community members who will be impacted by the outcomes. He also emphasizes that equity evaluations should focus on short-term impacts, since advocates tend to fight for urgently-needed improvements in the near future, and transit agency data model current conditions or short-term changes most accurately. This is a departure from industry standard, which prefers long-term modeling.

Remix, Remix Software Features to Measure Equity

Interviewee: Rachel Zack, former Director of Policy

Remix is a paid web-based software that displays spatial data, models scenarios, and explores equity implications of changes to transportation networks. With high usability and few skill requirements, Remix boasts analysis “with the click of a button” for its transit agency clients. The software features neighborhood demographics, import of any data source merged with a shapefile, an editor for existing GTFS feeds to reflect potential transit systems, and rapid computation.

Jane is a mapping feature that measures access to destinations on transit (and other modes). Using Jane, one can see how many destinations a transit rider can reach in an amount of time starting from an origin, and how long it takes to travel between two points. The user can modify the destination type (e.g. jobs, hospitals, schools, parks, etc.), the trip duration, and the network itself to evaluate how changing a transit stop, line, or network shifts access from neighborhoods.

Remix’s Title VI engine automates the required Title VI Service Equity Analysis. Users input proposed changes, and the engine estimates impacts on low-income and neighborhoods of color, in required report format. Going beyond Title VI, the engine weighs the potential change by how it shifts service frequency and by the number of people impacted by the change. The Title VI Engine simplifies what can be a days’ long analysis into a manageable, hours’ long task. In doing so, it benefits agencies, who won’t hesitate to change service for lack of capacity to complete the associated FTA requirement, and it can benefit riders if the changes improve service and are equitable. 

Remix does not advise users on how to interpret results from its software analysis. But it notes that the Title VI Engine analysis should only be a final check on equitable service — multiple equity measures and robust public engagement are necessary for an equitable system. As additional resources for that work, Remix proffers its Jane access to destinations metric and Explore, its spatial data layering tool. It also notes that software can be used to present tradeoffs of potential service changes, as a way to make community meetings more productive and interactive.