Equity Analysis in Regional Transportation Planning Processes, Volume 1: Guide (2020)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

A-1 A P P E N D I X Pilot Case Studies The outcome of the research conducted for TCRP Project H-54 is a succinct, readable resource guide that focuses on the key questions transportation agencies face when conducting equity analyses. During the early phases of the research, the team reviewed hundreds of online documents published by a nationwide sample of MPOs, followed by in-depth interviews with 10 agencies. Working with the advisory panel, the team selected four agencies with which to conduct pilot studies that would refine and enrich the contents of the guide: • Metro (Portland, Oregon): Developing key messages and communication vehicles for conveying results of an equity analysis to stakeholders and decision makers; • Denver Regional Council of Governments (Denver, Colorado): Testing the “population-weighted” approach for identifying populations to consider in equity analysis; • Mid-America Regional Council (Kansas City, Missouri/Kansas): Developing a strategic plan to engage a new regional equity network (REN) in the upcoming long-range transportation plan update, and (like Denver) testing the “population-weighted” approach for identifying populations to consider in equity analysis; and • Mid-Ohio Regional Planning Commission (Columbus, Ohio): Conducting GIS analysis to identify potential positive and negative impacts of new “smart city” transit investments. The objectives of the pilot project task were to ensure that the reference guide is a useful resource for practitioners and to benefit the agencies that are offering their time to help with pilot testing. Some of the pilot agencies are well-resourced, but we shaped the testing approach and documentation of each pilot to make the case studies useful to agencies with a wide array of capabilities, ideally allowing for replicability in many settings. The consultant project team considered issues and opportunities for implementing a few selected method(s) with each pilot agency and embedded those insights into the case studies as well as reflecting them in the main body of the guide. We strove to ensure that each pilot project case study was useful as an illustration of a given equity analysis method or approach, and that the tested method was contextualized within the broader planning and decision-making framework. Pilot Case Study Development Process 1. Initial discussions and assessments with staff. We identified key considerations and analysis methods for the agency to test and apply to its planning and decision-making process. For pilot testing to provide useful feedback, we applied the guidance to actual planning decisions. 2. Technical assistance from the project team over a period of several weeks, during which we continued reaching out to staff to offer help as needed. Some projects involved peer exchange conference calls and/or site visits to discuss methods and processes in depth. The content of the technical assistance varied based on the needs of each MPO, but primarily covered issues such as: a. Defining populations for analysis, b. Developing plans to engage key communities or stakeholders, c. Defining and identifying data sources for equity indicators, d. Conducting GIS analysis of indicators for different populations, and e. Facilitating meetings with stakeholders. 3. Case studies summarizing the interaction with each pilot tester agency.

A-2 Equity Analysis in Regional Transportation Planning Processes Metro (Portland, Oregon, MPO) Agency Context Metro, the MPO for the region surrounding Portland, Oregon, is a progressive organization which has taken many steps over the years to improve equity within transportation planning. The agency maintains an “Equity Dashboard” which tracks the racial and gender diversity of its workforce, and works closely with its Committee on Racial Equity, which is charged with overseeing and implementing a strategic plan to “remove barriers for people of color and improving equity outcomes for these communities by improving how Metro works internally and with partners around the Portland region.” In prior years, Metro conducted equity assessments primarily based on the amount of funding spent in underserved communities compared to other communities. For its 2018 Regional Transportation Plan, Metro tested new methods to assess the impacts of planned transportation investments in a more targeted and sophisticated manner and adjust their plan based on the assessment. Overview of Technical Assistance Metro staff discussed their outcomes-based approach to equity analyses on conference calls and one in-person meeting with the project team in order to inform the project team of their process in the context of this guide and to identify options for communicating both the process and the results of the process. The objective of this pilot was to observe and discuss Metro’s outcomes-based assessment of transportation investments, which measured how projects would impact underserved persons in a meaningful and understandable way. The project team provided technical assistance to identify key messages and an appropriate communication vehicle to describe the results of the assessment. Messages would be suitable for elected officials, interest groups, and the general public. Method Metro conducted an equity evaluation of proposed transportation projects using a set of measures indicating a variety of impacts on underserved persons. Underserved communities were defined as communities meeting threshold percentages of persons of color, persons in poverty, persons with Limited English Proficiency (LEP), older adults, and young people. The following measures were selected as representing priorities for the Regional Transportation Plan (RTP) identified during the public input process: • Access to jobs: Number of low/medium/high wage jobs accessible by a “reasonable” commute time via walking, bicycling, transit, and driving. Commute times defined as 20 minutes by walking, 30 minutes by bicycling, 45 minutes by transit, and 30 minutes by automobile. Data used: Geospatial project information for proposed transportation projects provided by project sponsors; forecast employment/jobs determined by Metro’s Metroscope Model. • Access to community places: Number of community places, such as schools, religious organizations, libraries, health services, and so forth, accessible by a “reasonable” commute Pilot at a Glance Region: Portland, Oregon MPO Service Population: ~1,500,000 Phase: Long-Range Transportation Planning, Programming Focus of Technical Assistance: Assess Relative Impacts

Pilot Case Studies A-3 time via walking, bicycling, transit, and driving. Data used: Geospatial project information for proposed transportation projects from project sponsors; U.S. Bureau of Labor Statistics – Quarterly Census of Employment and Wages (2013); and North American Industry Classification System (NAICS) codes. • Share of safety projects: Number and percentage of transportation safety projects compared to total RTP investment packages; percentage of total cost of investment strategies; percentage of transportation safety investments per capita regionwide, in underserved communities, and on high injury corridors. Data used: Geospatial project information for the investment strategies from project sponsors; location of high injury corridors identified by Metro, the Oregon Department of Transportation (DOT) Pedestrian and Bicycle Safety Implementation Plan, state/city/county safety plans, high injury intersections, and areas with one or more severe crashes in the previous 5 years. • Exposure to severe crash risk: Change in non-freeway average vehicle-miles traveled. Data used: Geospatial project information for proposed transportation projects provided by jurisdiction; forecast vehicle-miles traveled by TAZ from the travel-demand model. • High value habitat impact: Percentage of transportation projects intersecting identified resource habitats. Data used: Geospatial project information for the investment strategies from project sponsors; geospatial resource conservation information from Metro-identified resource and conservation habitat areas. The measures were estimated for a constrained scenario (2040), strategic scenario (2040), short- term constrained scenario (2027), and no-build scenarios for both 2027 and 2040. After conducting the equity analysis for submitted project list, Metro published the results and called for a second round of project submissions from its local partners, providing an opportunity to revise their list of projects to better serve underserved communities. After receiving the second list of projects from partners, Metro conducted another equity analysis. The second analysis used the same measures, but an improved methodology. The two-round call for projects was a new process for the 2018 RTP. At this stage, Metro planned to communicate the results of the equity analysis to elected officials, partners, stakeholders, and the general public, both to describe the efforts undertaken to analyze the impacts of transportation investment on the region as a whole as well as specifically on underserved communities, and also to illustrate the work still required to address discrepancies between underserved persons and other populations. Figure A-1 summarizes underserved communities defined in Metro’s 2018 RTP Transportation Equity Evaluation.

A-4 Equity Analysis in Regional Transportation Planning Processes Figure A-1. Underserved communities, defined in the Portland [Oregon] Metro 2018 RTP Transportation Equity Evaluation. Results After discussions during the pilot, Metro decided to focus communication efforts on the access to jobs measure. The ability to access jobs within a reasonable commute is a clear and common objective for many people and can highlight the relevance of transportation planning on the everyday lives of Metro’s citizens. The proposed transportation investments were shown to increase access to jobs across all wage levels for all communities in the region, however, underserved communities were predicted to have a smaller increase than other communities. Among all modes, transit improvements were the most substantial. For all communities, the percent increase in access to jobs via transit was over four times the percent increase in access to jobs by any other mode. As underserved communities may be more reliant on transit than the general population (related to income discrepancies or abilities), this demonstrated investment in transit may be interpreted as a positive strategy for the RTP. There are other potential explanations for the discrepancy in increased access to jobs, other than an imbalance in transportation investment. For example, increased traffic congestion or changes in land use may impact commute times, decreasing the number of jobs accessible within the model parameters of a “reasonable commute.” Despite other possible explanations, the results of the analysis indicate that Metro as a region should make strategic investments to support transportation for persons living in underserved communities. The project team drafted simple infographics (Figure A-2) illustrating the results of the measure, showing that while all people living in the region will have access to more jobs, those living in underserved communities will have a smaller percentage increase than others.

Pilot Case Studies A-5 Source: Metro, 2018 Regional Transportation Plan, Transportation Equity Evaluation Appendix. Figure A-2. Draft infographics illustrating discrepancies in access to jobs by mode between historically marginalized communities and non-historically marginalized communities.

A-6 Equity Analysis in Regional Transportation Planning Processes Lessons Learned: Implications for the Equity Guide Metro’s equity analysis assesses specific impacts to underserved communities in a way that is understandable and meaningful. Stakeholders can understand how the investment will make a difference in how the transportation system will connect them to jobs, key destinations, and help improve safety on the streets. Offering a second call for projects to local partners after conducting an initial equity analysis provides an opportunity for partners to reflect and re-evaluate their project lists. As the analysis methods evolved over the development of the 2018 RTP, Metro was unable to compare the change in equity impacts between the two rounds. Ideally, the methodology for the analysis would be identical between the two rounds in order to compare potential improvements. Illustrating the discrepancies in job access is an important step in addressing inequalities. This discrepancy can be seen as a call to action for local partners in focusing a greater share of their transportation investment on projects that will help underserved persons access jobs and other destinations. Resources • Metro. 2018. Regional Transportation Plan, Transportation Equity Evaluation Appendix. Retrieved from: https://www.oregonmetro.gov/sites/default/files/2018/06/29/RTP-Appendix_E_2018_RTP_ Transportation_Equity_Evaluation_with_attachments.pdf.

Pilot Case Studies A-7 Denver Regional Council of Governments (DRCOG) Agency Context The DRCOG is a large MPO serving the Denver metropolitan region, which includes adjacent cities and rural areas in Colorado. DRCOG was included in the research for this guide because the agency has a strong working relationship with the area’s transit agency (Regional Transit District) and is often held up as an example of good practices for public engagement and equity analyses. The literature review and subsequent interviews revealed that DRCOG is also one of relatively few MPOs that have used their travel- demand models as part of their equity analyses. On reaching out for this study, the research team learned that DRCOG has been dissatisfied by the approaches for designating certain areas as equity areas on the basis of some threshold concentration levels of underserved persons. DRCOG staff have been looking for options for expanding their equity analyses beyond solely comparing geographic areas that meet the thresholds to areas that do not meet the thresholds. During the course of the technical assistance and pilot, DRCOG and the technical support team discussed alternative approaches to the process of identifying equity areas, and the agency experimented with the population-weighted approaches used by the Mid-Ohio Regional Planning Commission (MORPC). The pilot also briefly covered options for getting equity focus groups to weigh in on impacts and options for evaluating whether TIP projects improve equity; given the timeframe of the pilot, DRCOG opted not to further explore these options at this time. Overview of Technical Assistance The objective of this pilot was to identify how the approaches suggested in the draft guide could help DRCOG refine its approach to equity analyses to avoid issues that DRCOG had identified with its current process, in which it identified areas as equity or non-equity areas on the basis of the area’s percentage of underserved persons. Two issues in particular with the all-or-nothing approach to defining an “environmental justice zone” (EJ zone) were as follows: • More than half of the region’s population lives in an EJ zone. DRCOG would like to improve its ability to focus strategically on high-priority zones. • Significant numbers of underserved persons live outside the designated EJ zones. DRCOG would like to ensure that its analyses capture impacts to those underserved persons. DRCOG also would like to develop quantitative methods to improve their current qualitative approaches for assessing whether projects in their plans or programs have a disproportionate adverse effect on EJ populations. Method Tested DRCOG and the technical support team held a series of calls to discuss potential approaches. Each discussion built on prior calls to further refine the mutual understanding of the challenges Pilot at a Glance Region: Denver, Colorado MPO Service Population: 2,827,082 Phase: Long-Range Transportation Planning, Programming Focus of Technical Assistance: • Identifying Populations, • Measuring Impacts, • Population-Weighted Analysis

A-8 Equity Analysis in Regional Transportation Planning Processes with thresholds-based approaches and to explore the population-weighted approach as an alternative. DRCOG staff reviewed the population-weighted approach described in the draft guide, and the technical support team answered questions DRCOG had about implementing the approach. The technical support team organized and facilitated a peer-to-peer web-enabled call with staff from MORPC, in which DRCOG (and MARC) staff asked questions they had about the approach. • Identifying populations for analysis: Rather than using a threshold approach, MORPC uses a combination of heat maps showing relative percentages and a dot-density map showing density numbers. • Determining disproportionate adverse effect: MORPC uses line graphs showing how the modeled benefits (such as travel-time savings) vary by population group for the current/base year, the no-build in the forecast year, and the plan/program’s projects in the forecast year. These line graphs enable MORPC staff to identify gaps and to see how improvements might vary for different populations. • Long-term applicability of Census Transportation Planning Products (CTPP): The CTPP will be phasing out TAZs in favor of block groups, which diminishes the usefulness of the CTPP. MORPC used census products early on but has since created their own population synthesizer (using Urban Sim population data) to calculate population demographics of TAZs. Other agencies could start with the CTPP and then develop their own approaches in later years. DRCOG already has the demographic data for their TAZs but had previously only used that data for comparisons between EJ zones and non-EJ zones; the same data can be used to develop a population-weighted approach. Results DRCOG staff tested the use of the population-weighted approach in the fall of 2018 as the agency prepared for its long-range transportation plan update. The test asked the following questions: • What MORPC EJ measures are replicable in the Denver region? • What EJ measures should be added for the Denver region? • What data are necessary for the analysis; is that data available in the Denver region? • How capable is DRCOG of replicating the analysis from start to finish? Analyses were conducted on four selected measures: 1. Target/Non-Target by Zones, 2. Average Number of Employers Close (within 20 minutes for auto and 40 minutes for transit), 3. Average Travel Time for All Purposes, and 4. Average Travel Time to CBD [Central Business District]. The test proved that it was possible for DRCOG to replicate some measures of the MORPC EJ technical analysis. Next steps include combining and weighting the information by population group. It may also be necessary for the DRCOG transportation team and information systems team to brainstorm several more measures before finalizing this work.

Pilot Case Studies A-9 Analysis Process Identification of Replicable Measures EJ measures “. . . compare the relative treatment of the target populations and non-target populations in the planning process.” The MORPC methodology covered the following measures: • Target/Non-Target by Zones, • Average Number of Job Opportunities Close, • Average Number of Shopping Opportunities Close, • Average Number of Non-Shopping Opportunities Close, • Percent of Population Close to a College, • Percent of Population Close to a Hospital, • Percent of Population Close to a Major Retail Destination, • Average Travel Time for Mandatory Purposes, • Average Travel Time for Shopping Purposes, • Average Travel Time for Other Purposes, • Average Travel Time for All Purposes, • Average Travel Time to CBD, • Transit Access to CBD, • Congested Vehicle-Miles of Travel during Peak Hours, • Transportation Investments, and • Displacements from Highway Projects. Of these 16 MORPC measures, DRCOG identified the following eight as immediately replicable based on current data in the Denver region (strikeouts indicate where DRCOG data varies slightly from MORPC data): • Target/Non-Target by Zones, • Average number of [Job Opportunities] Employers Close, • Percent of Population Close to a College, • Percent of Population Close to a Hospital, • Average Travel Time for [Mandatory] Job Purposes, • Average Travel Time for All Purposes, • Average Travel Time to CBD, and • Transit Access to CBD. Additional Measures Identified During the preliminary assessment of EJ measures used by MORPC, DRCOG identified the following additional measures as potential for study in the Denver region: • Average number of Grocery Stores Close, • Average travel time to VA Medical Facilities, • Average travel time to Public Recreation Centers, and • Transit Access to all Regional Employment Centers. More measures may be created or identified if a full DRCOG EJ technical analysis moves forward.

A-10 Equity Analysis in Regional Transportation Planning Processes Data Required for Methodology creates future horizon-year estimates including data for the year 2040. In the MORPC analysis, data for two different 2040 forecast scenarios were presented for each measure (first, forecast travel assuming no growth in transportation system, second, forecast travel with growth in the transportation system). DRCOG represents data for only one 2040 scenario, using projections that assume growth in the transportation system. Travel model data is aggregated at TAZs for off-peak times and peak times, and for automobiles and transit. This results in eight total scenarios used from the travel model—Off-Peak Automobile 2015, Off-Peak Transit 2015, Peak Automobile 2015, Peak Transit 2015, Off-Peak Automobile 2040, Off-Peak Transit 2040, Peak Automobile 2040, and Peak Transit 2040. Measurements Identified for Test In the interest of time, this test application of the MORPC methodology addressed only the following measures: • Target/Non-Target by Zones, • Average Number of Employers Close, • Average Travel Time for All Purposes, and • Average Travel Time to CBD. Test Measurement Methodology Data analysis occurred for the four test measures identified. The following sections describe the technical details of analysis by measure. Target/Non-Target by Zones As stated in the MORPC analysis, “In order to create the population-based measures, it is necessary to estimate the target and non-target population within each TAZ. However . . . only Figure A-3. Example illustration of ACS 2016 5-year data: percentage minority for TAZs in the Denver region. Most of the measures described by MORPC in their analysis rely on Census ACS data (see Figure A-3). The Census ACS data provides information on target populations—Disabled, Hispanic or Latino, In Poverty, Minority, Over 65, Zero-Car Households—by block group and tract. Many measures also rely on the DRCOG travel forecasting model process. The travel forecasting model process takes land use and transportation information and estimates travel times, patterns, and volumes on the transportation system. The most current transportation modeling data for DRCOG is 2015, and DRCOG also

Pilot Case Studies A-11 total population by TAZ is developed.” At DRCOG, target and non-target population information from 2016 Census ACS data is summarized by census block group and census tract and distributed publicly as a “Vulnerable Populations” dataset. (Despite the 1-year difference from travel model work, this data was used for the purposes of this methodology test. It is also significant to note that because of this data source, this test only covers the DRCOG planning region rather than the entire transportation modeling region.) In order to apply target population information to TAZ geographies it was first necessary to create a Tract-to-TAZ equivalency table and a Block-Group-to-TAZ equivalency table using weighted percentages based on population and housing. To build the equivalency tables, DRCOG used census block centerpoints to determine approximately what percentage of each Tract and Block Group population or housing stock fall inside of each TAZ. Census/TAZ overlap geometries were defined as new polygon “equivalency” layers. Using this join method for every census/TAZ overlap, the total population of blocks within each “equivalency” overlap polygon was determined. This number was then compared to the Tract or Block Group population total to define the approximate percentage of each block group or tract that is represented within each TAZ (see Figure A-4). In Figure A-4, TAZs are represented by black lines, and block groups are represented by red lines. Block centerpoints are then used to create an equivalency layer (shown in blue in Figure A-4). Block totals were also compared to Tract and Block-Group totals to quality check the block information. Using the equivalency layer percentages, target population data from 2016 Census ACS data was added to TAZ geographies. Then, another group of percentages was calculated to weight forecast 2040 populations by assuming the same percentage totals. The resulting data thus includes target population information by zone for both different scenarios—2015 and 2040—and Figure A-4. Illustration of block group- to-TAZ equivalency.

A-12 Equity Analysis in Regional Transportation Planning Processes for all six target populations (Disabled, Hispanic or Latino, In Poverty, Minority, Over 65, Zero- Car Households). From MORPC, “In estimating the target populations by traffic zone, it was assumed that [in 2040] the total regional percentage for each population would be the same percentage as the 2015 percentage. For example, the regional percentage in poverty in 2015 was 13.9%. Thus, for the forecast 2040 populations, it was assumed that the regional poverty percentage would remain at 13.9%.” Average Number of Employers Close This measure nearly replicates the MORPC measure “Average Number of Job Opportunities Close” using the DRCOG’s master employment dataset. The measure estimates the average number of employers within a specified travel time from each zone. • The raw number of employers by TAZ was first calculated using a simple spatial join. • Next, information from the DRCOG travel model was used to estimate peak and non-peak period auto and transit travel times from each TAZ to every other TAZ (a travel-time skim). • For each TAZ based on the skim, the total number of employers located in TAZs within 20 minutes by auto (see Figure A-5) and 40 minutes by transit were calculated. • A weighted average of the number of employers can be calculated based on the number of each target population group within each TAZ. In Figure A-5 (next page), the top map shows the number of employers within 20 minutes for 2015 non-peak skims, and the bottom map shows the number of employers within 20 minutes for 2040 peak skims. Average Travel Time for All Purposes As described in the MORPC documentation, “through the modeling process, different tour purposes are defined . . . . The previous measures were accessibility measures. This measure, however, is a travel estimate measure.” To calculate average travel times for all purposes from each TAZ, the TAZ-to-TAZ baseline totals were summed by the originating TAZ and an average travel time was calculated (see Figure A-6). In Figure A-6, the top map shows average travel times for 2015 non-peak skims, and the bottom map shows average travel times for 2040 peak skims. As the MORPC points out, “exact population groups using the different modes is unknown. Thus, when calculating the measure for a particular mode, the weighted average is based on the proportion to the target and non-target population in the origin zone . . . .” Next, the weighted average of travel time by population group can be calculated.

Pilot Case Studies A-13 Figure A-5. Example illustration of employers within 20 minutes for auto skims.

A-14 Equity Analysis in Regional Transportation Planning Processes Average Travel Time to CBD At DRCOG, significant urban centers are named and distributed publicly as an “Urban Centers” dataset that helps guide planning efforts. This measure pulled the average travel time from each TAZ in the region to the single TAZ containing the DRCOG “Central Business District” urban center centroid. Next, the weighted average travel time by population group can be calculated. Figure A-6. Example illustration of travel times to all TAZs for auto skims.

Pilot Case Studies A-15 In Figure A-7, the top map shows average travel times to CBD for 2015 non-peak skims, and the bottom map shows average travel times to CBD for 2040 peak skims. Figure A-7. Example illustration of travel times to CBD for auto skims.

A-16 Equity Analysis in Regional Transportation Planning Processes Lessons Learned: Implications for the Equity Guide Based on the questions posed by DRCOG and the conversation held during the peer-to-peer call, the technical support team revised the relevant sections of this guide to provide additional clarity and explanation. Resources • DRCOG. Equity Atlas. Available at: http://www.denverregionalequityatlas.org. • DRCOG. 2017. 2040 Metro Vision Regional Transportation Plan. Retrieved from: https://drcog.org/sites/default/files/resources/FINAL%20-%202040%20MVRTP%20w%20APPENDICES%20- %20April%202017.pdf. • DRCOG. 2010 Public Participation Plan. Retrieved from: https://drcog.org/sites/drcog/files/resources/FINAL%20DRCOG%20Public%20Involvement%20in%20Regional %20Transportation%20Planning%20Adopted%20April%202010.pdf.

Pilot Case Studies A-17 Mid-America Regional Council (MARC) Agency Context The Mid-America Regional Council (MARC) serves as the MPO for the Kansas City region. The MPO spans nine counties in two states. MARC is committed to advancing racial and economic equity. In 2010, MARC established the Creating Sustainable Places Initiative, with funding from the U.S. Department of Housing and Urban Development. The initiative set out to create a regional vision for sustainable development, with a focus on green, healthy, and vibrant places offering a range of housing options, amenities, and services, that are well connected by multiple transportation options. The Creating Sustainable Places Initiative established equity as a core principle for regional planning and decision making. MARC also developed an equity lens to facilitate a multi-dimensional consideration of equity that addresses education, public engagement, EJ, housing choices, transportation, health, and reinvestment. Overview of Technical Assistance MARC received technical assistance related to public engagement and assessing relative impacts for the development of its long-range Regional Transportation Plan 2050. Specifically, the technical assistance included (1) guidance on conducting a population-based equity analysis, and (2) the development of an engagement plan to involve the existing REN (a partnership of local organizations formed to support the Creating Sustainable Places Initiative) in the public engagement process. Method Population-Based Analysis Method Guidance MARC piloted the draft guide’s instructions for conducting a population-weighted equity analysis, which forecasts the outcomes of transportation investments on minority persons and low- income persons throughout the region. A population-based approach to an equity analysis helps reveal impacts to all underserved populations, not just those living in high-priority underserved communities. The population-based equity analysis could replace the geographic-based analysis used in previous regional transportation plans. To support the implementation of this new approach, MARC received a draft version of the reference guide that documents the population-based approach used by the MORPC. MARC successfully used the approach by following the draft guide’s instructions but found that the agency’s regional travel-demand model did not forecast transit-based measures as accurately as they would like. The technical support team organized and facilitated a peer-to-peer webconference with staff from MORPC, in which MARC (and DRCOG) staff asked questions they had about the approach. Pilot at a Glance Region: Kansas City, Kansas/ Missouri MPO Service Population: 2,086,771 Phase: Long-Range Transportation Planning Focus of Technical Assistance: • Public Engagement, • Use of Population-Weighted Approach for Assessing Relative Impacts

A-18 Equity Analysis in Regional Transportation Planning Processes • Defining populations for analysis: MORPC always conducts separate analyses for each population group, rather than combining minority populations with low-income populations. MORPC always includes a control group to enable comparative analyses of the results; for example, when they analyze minority populations, they also analyze non-minority populations. • Identifying populations for analysis: MARC staff had questions about developing the forecast year’s population figures. – MORPC explained that they needed to hold three things constant: (1) the regional demographics held constant with the base year’s; (2) the regional overall population number for the forecast year; and (3) the total population of each TAZ in the forecast year. – To successfully hold those three things constant while accommodating the different growth rates forecast for different TAZs, they had to redistribute slightly at the TAZ level but maintained each TAZ’s relative share of the regional population for each population group being analyzed. – They identified which population groups had declined at the regional level and how far they were off. Then they adjusted TAZs based on each TAZ’s current share of the region’s population of that population group. For example, if they needed to add 10,000 more low-income persons to the region, then they would distribute an additional 10,000 low- income persons to the TAZs based on the current proportions of low-income persons in each TAZ, and then reduce each TAZ’s non-low-income numbers by the same amount to hold the TAZs’ total population constant. • Identifying high-priority areas: MORPC uses a combination of heat maps showing which TAZs have high concentrations of each population group and dot-density maps showing where high numbers of each population group live. • Determining disparate impacts and/or disproportionately high and adverse effects (DHAE): MORPC uses line graphs showing how the modeled benefits (e.g., travel-time savings) vary by population group for the current/base year, the no-build in the forecast year, and the plan/program’s projects in the forecast year. These line graphs enable MORPC staff to identify gaps and to see how improvements might vary for different populations. • Identifying mitigation strategies: Although MORPC did not identify disparate impacts or DHAE, MARC was interested in how the use of this approach influenced decision making at the agency. MORPC then described its approach of requiring project sponsors to forecast the opening-day users of proposed projects, which MORPC compares to the region’s demographics to ensure that they are adequately funding projects that benefit a representative sample of the region. REN Engagement Plan The REN Engagement Plan identifies opportunities for targeted engagement of the REN members in the development of the long-range transportation plan. Currently, the REN participates in MARC’s transportation committee and provides input to transportation planners in one-on-one meetings. The engagement plan helps to cultivate dual roles for the REN member organizations to (1) serve as a focus group to provide input on the transportation needs, concerns, and ideas from the perspective of underserved persons; and (2) serve as engagement ambassadors to support MARC’s efforts to effectively engage with underserved persons.

Pilot Case Studies A-19 The purpose of the plan is to identify opportunities for MARC to meaningfully engage members of the REN in the development of the long-range Regional Transportation Plan 2050. Engaging REN members will support the understanding and consideration of transportation needs, concerns and ideas from the perspective of underserved populations, helping to ensure that regional transportation planning is more inclusive and equitable. About the REN In 2010, MARC received a Sustainable Communities Planning Grant from the U.S. Department of Housing and Urban Development. With the grant, MARC established the Creating Sustainable Places Initiative to make the region more “vibrant, connected, and green.” A set of guiding principles were developed, which included the following definition of equity (MARC): Residents of all races, economic means, and abilities are welcome and equipped to participate in all aspects of community life. A region is most likely to be sustainable, and nationally and globally competitive, if all its residents are active participants in its economy, community, and public life. As part of the initiative, an “equity lens” was developed that identifies potential strategies, programs, and policies to address equity. The lens provides a multi-dimensional consideration of equity that addresses education, public engagement, EJ, housing choices, transportation, health, and reinvestment. One of the first applications of the equity lens was the evaluation of six corridor- planning demonstration projects (MARC). To support the Creating Sustainable Places Initiative, local organizations formed the REN to advance the consideration of equity. The group’s mission is to “ensure that all planning processes and policy decisions take social equity priorities into account” (MARC). REN Members • Ad Hoc Group Against Crime, • Blue Hills Housing Development Corporation, • Communities Creating Opportunities, • Greater Kansas City Local Initiative Support Corporation, • Health Department of the City of Kansas City, • Ivanhoe Neighborhood Council, • Latino Civic Engagement Collaborative, • MORE2 (Metro Organization for Racial and Economic Equity), • MARC, • Neighborhood Housing Services Department, City of Kansas City, • Northland Neighborhoods, Inc., • University of Missouri Kansas City, Urban Planning Department, • Upper Room, • Urban League, • William Jewell College, and • Westside Housing community leaders, stakeholders, and individual members. Regional Equity Profile The REN’s first major activity was to develop a Regional Equity Profile to better understand the state of social equity in the nine-county area served by MARC. REN worked with PolicyLink and

A-20 Equity Analysis in Regional Transportation Planning Processes the Program for Environmental and Regional Equity at the University of Southern California on an analysis of the composition and spatial distribution of minority populations and low-income residents in the region (Figure A-8). The profile presented an equity indicators framework with a focus on four key areas: demographics, economic vitality, economic readiness, and connectedness to regional assets and opportunities. The measures were developed using data from a regional equity database as well as public and private data sources. The analysis found growing inequalities related to employment, income, and education despite steady regional economic growth. The Regional Equity Profile recommends that, “given the region’s rapid demographic shifts, public-sector leaders need to take steps to ensure active and accessible public engagement by all of its racial and ethnic communities in local and regional planning processes” (http://www.marc.org/Regional-Planning/Creating-Sustainable-Places/Plans/Social-Equity). The REN Engagement Plan helps accomplish this objective by identifying concrete strategies for meaningful engagement from communities of color and low-income residents in the update of the RTP. Specific insights gleaned from the profile that were considered in the development of the REN Engagement Plan are summarized in Table A-1. Source: PolicyLink and U.S.C. Program for Environmental & Regional Equity (2013), Equity Profile of the Kansas City Region Figure A-8. Concentrations of households in poverty and people of color.

Pilot Case Studies A-21 Table A-1. Equity profile insights for REN Engagement Plan. Regional Equity Profile Findings Insights for REN Engagement Plan The Kansas City region is rapidly diversifying in the urban core and suburbs. Communities of color have contributed to most of the region’s recent population growth, particularly among youth. The Latino community is almost evenly comprised of U.S.-born and foreign-born individuals. The Asian community is very diverse, representing people with roots in nations such as China, Taiwan, Viet Nam, the Philippines, and Korea. • Identify strategies to encourage participation of both U.S.-born and immigrant Latinos, which may include addressing the language barrier. • Acknowledge the diversity of the Asian community and seek out participation from the various cultural groups. • Make concerted efforts to engage people of color throughout the region. Communities of color, especially youth, face educational gaps and health challenges. Black and Latino communities lack the education and training demanded by employers and job forecasts. Also, the prevalence of disease, such as obesity, diabetes, and asthma is higher, in part due to lack of access to medical care. • Make concerted efforts to engage youth within communities of color. • Consider how transportation decision making impacts health (e.g., access to medical care and opportunities for walking and biking). Although residential segregation is decreasing overall, areas of concentrated poverty present a growing challenge and people of color are more likely to reside in these areas. • Consider the compounding impact of multiple layers of disadvantage. • Conduct targeted engagement in areas with intersecting high concentrations of poverty and of people of color. Regional Transportation Plan 2050 Timeline and Engagement The development of the Regional Transportation Plan 2050 is a six-step process (Figure A-9). MARC completed the first two phases of the process prior to the pilot study: Discovery and Needs Assessment and Storytelling and Policy Framework. The next phase, Investment Scenarios, was slated to begin in the fall of 2018. Figure A-9. MARC Regional Transportation Plan 2050 development schedule. Public and stakeholder engagement is a critical component of MARC’s planning process. The Regional Transportation Plan 2050: Public & Stakeholder Engagement Plan details the underlying principles and strategies for the public engagement process: • Active and continuous process: MARC is committed to fostering a public participation process that provides equal access to engagement opportunities. MARC strives to facilitate

A-22 Equity Analysis in Regional Transportation Planning Processes meaningful public engagement, in which members of the public understand how their input influences transportation decision making. • Inclusive engagement: Efforts to address equity include the participation of the REN in the RTP steering workgroup to provide guidance on the planning process and to encourage engagement among advocacy and faith-based organizations. Social service agencies serve as a voice for transportation-disadvantaged populations such as older adults, youth, persons with disabilities, persons with LEP, ethnic and racial minority community members, and low-income households. These groups participate in MARC’s transportation committee and provide input during one-on-one meetings. To effectively engage this group, MARC will provide education about the decision-making process and facilitate dialogues with members of low-income and minority communities and their representatives. Furthermore, MARC will document input from transportation-disadvantaged groups for review and consideration by decision makers, and communicate back to these groups about how their feedback is being addressed in the plan. • Broad communications and engagement strategies: MARC uses a wide range of strategies and tools to communicate, educate, and engage, including a website, social media, press releases, newsletters, videos, story maps, presentations, targeted engagement, one-on-one meetings, “piggyback” presentations, public meetings, open houses, pop-up meetings, online surveys, workshops, contests, youth-focused engagement, participation in regional events, and advertisements in print, online, and radio media. • Evaluation: The plan states that MARC will use evaluation metrics (quantitative and qualitative) to measure the effectiveness of engagement techniques. These may include, for example, numbers of participants, completed surveys, geographic distribution and demographics of participants (age, income, gender, race, ethnicity, and so forth), and web analytics. Action Strategies for REN Engagement The REN Engagement Plan makes recommendations to MARC on how to meaningfully engage REN member organizations in the update of the RTP. There are two key roles that REN members can serve in the public engagement process: 1. Serve as a focus group to convey to planners the transportation needs, concerns and ideas from the perspective of transportation-disadvantaged groups in the region and provide technical expertise; and 2. Serve as engagement ambassadors to support MARC’s efforts to meaningfully and effectively engage with transportation-disadvantaged groups. Due to the timing of the pilot project, the recommendations in this plan identify preliminary activities that would enable REN members to provide input on completed phases, as well as strategies for expanding REN engagement in the remaining phases. Preliminary Activities The first two phases of the plan development process gather information about big-picture needs in the region and shape the story of transportation in the region—present and future. These steps create the foundation on which the remaining phases—and ultimately the long-range plan—will be built. Therefore, it is important that the narrative and policy framework developed during these phases depict an accurate picture of the current transportation system and envision a more equitable

Pilot Case Studies A-23 future. The following preliminary activities are recommended to enable REN members to validate the outcomes of the previous steps and to set the stage for meaningful engagement throughout the remainder of the process. • Examine the current REN roster and suggest additional members to ensure that it reflects the diversity of the region found in the Regional Equity Profile. Specifically, REN should also include groups that work on behalf of youth, both U.S.- and foreign-born Latino immigrants, and the various Asian cultures found in the region. • Convene REN members to review and provide feedback on the needs assessment, regional story narrative, and policy framework developed in the earlier phases. REN members can confirm whether the needs of low-income populations and communities of color have been accurately identified. • Work with REN members to develop measures of effectiveness for communication, education, and engagement strategies that foster an inclusive public and stakeholder engagement (such as participants’ demographic information). Consider evaluation goals and targets, as well as methods for documenting, tracking, and responding to evaluation results. • Seek input from REN members on which public engagement techniques might be most effective for reaching various audiences (such as interactive games, group discussions, one- on-one interviews, and surveys). Also ask them to identify places to hold workshops and other activities and distribute meeting advertisements for targeted engagement to underserved communities. • Provide materials for REN members to share the story map and policy framework through social media, agency websites and “piggyback presentations.” This can help set the stage for targeted engagement of transportation-disadvantaged populations in subsequent phases. Action Strategies for Remaining Phases of Plan Development These preliminary activities will lay the foundation for the REN members to serve as both technical consultants and engagement partners in the subsequent phases to begin in the fall of 2018. Table A-2 lists suggested strategies for REN engagement in each of the remaining plan development phases. These recommendations include a broad array of activities that MARC can use to engage REN members in the long-range planning process, with varying levels of time, resource, and capacity requirements. As a first step toward implementing these or other ideas, MARC will gauge REN members’ interest in these potential strategies, which could lead to the development of alternative and/or additional activities Lessons Learned: Implications for the Equity Guide • Public engagement: Involve local stakeholder groups early in the planning process, and develop strategies to leverage their availability and capacity most effectively, especially given time and resource constraints of stakeholders operating as nonprofits. • Population-based method: Conducting a population-based equity analysis is an effective tool in understanding impacts on underserved individuals throughout a region. The technical analysis is not significantly more labor intensive than a geographic-based approach. By using a population-based analysis, project sponsors can be required to include projected numbers or percentages of a project’s users, broken down by demographics. Projects serving higher

A-24 Equity Analysis in Regional Transportation Planning Processes percentages or numbers of individuals from underserved groups can be prioritized in project programming. Guidance on how to analyze transit travel-time trends would be helpful. Table A-2. Potential strategies for REN engagement in Regional Transportation Plan 2050. Phase: Investment Scenarios Objective: Develop alternative scenarios for regional transportation investments. REN Role: Help shape scenarios that address the needs of transportation-disadvantaged populations. Potential Action Strategies: • Educate REN members about scenario planning, particularly because this is the first time scenario planning is being used to develop the long-range transportation plan. • Produce a brief video featuring REN members to educate the public—particularly transportation- disadvantaged groups—about scenario planning and transportation decision making to increase interest in public engagement activities. The video could be shared online and played at public engagement events. Example: Memphis MPO YouTube Channel video https://www.youtube.com/user/memphismpo. • Create opportunities for REN members to serve as engagement ambassadors to draw transportation- disadvantaged groups to participate in regional stakeholder workshops and online survey/poll. Focus on areas with high concentrations of low-income residents and minority populations. • Seek input from REN members on quantifiable indicators to incorporate the equity lens to the selection of the preferred scenario. Work with REN members to determine measures of effectiveness for policies included in the policy framework. Establish guiding principles for the scenario analysis such that the preferred scenario should not harm low-income communities and communities of color. • Leverage REN interest in housing to consider the linkages between housing and transportation. Utilize the Housing + Transportation Index and other data on housing availability and affordability to enrich scenarios to consider land use impacts as well. Also consider data on health to address how scenarios might help address health disparities identified in the Regional Equity Profile. • Ask REN member organizations to invite MARC to existing community events/meetings to present on the planning process and capture input and ideas for consideration. • Provide training to REN members to facilitate activities during public workshops and meetings. • Hold activities to specifically involve youth. For example, hold charrettes for teenagers, use schools as publ ic meeting spaces, or partner with youth organizations (e.g., Boys and Girls Club or youth organizations in Latino or Asian communities). Engaging youth can also be an effective strategy for involving their parents/guardians in the public engagement process. • Ask REN member organizations to consider contributing culturally-appropriate incentives to encourage participation from underserved groups (e.g., catering, translation services, raffle prizes, and transit passes). Phase: Project Selection Objective: Identify list of projects to accomplish transportation vision. REN Role: Support engagement of transportation-disadvantaged in the selection of projects that address transportation equity. Potential Action Strategies: • Seek input from REN members on the project selection criteria to incorporate the equity lens. Equity-related criteria might include, for example, access to jobs and services, transportation choices, and household transportation costs. Connect project selection criteria back to the preferred scenario. • Ask REN for input on projects with a higher potential for burdens and benefits impacting underserved populations. • Develop specific opportunities for REN members to serve as engagement ambassadors (time and resources permitting) during regional workshops and open houses. For example, train them to be discussion facilitators, activity leads, greeters, and scribes. • Develop communications resources (e.g., sample messages, boilerplate text, images, and graphics) for REN members to share opportunities for public comment on individual projects through their social media, newsletters, websites, and other outlets. (continued on next page)

Pilot Case Studies A-25 Table A-2. (Continued). Phase: Plan Development Objective: Prepare long-range transportation plan. REN Role: Support engagement of transportation-disadvantaged in the selection of strategies that address transportation equity. Potential Action Strategies: • Leverage the expertise of REN members and seek their input on draft strategies and plan narrative. • Work with REN members to identify and address any equity-related impacts (positive and negative) of plan recommendations. • Develop communications resources (e.g., sample messages, boilerplate text, images, and graphics) for REN members to share opportunities for public comment on the draft plan through their social media, newsletters, websites, and other outlets. • Develop specific opportunities for REN members to serve as engagement ambassadors (time and resources permitting) during regional workshops, pop-ups, and open houses. For example, train them to be discussion facilitators, activity leads, greeters, and scribes. • Coordinate with REN members to hold engagement activities in areas with high concentrations of low-income residents and minority populations. • Communicate to REN members how their input—as well as input from low-income communities and minority populations—is being considered in the plan. • Collaborate with REN members to identify opportunities for alignment between plan recommendations and agency programs and initiatives. Leveraging transportation investments with REN activities helps to maximize benefits, promote integrated planning, and advance smart investments. Phase: Plan Adoption Objective: Adopt final long-range transportation plan. REN Role: Help communicate to transportation-disadvantaged groups how their feedback was considered and addressed in the final plan, and encourage their continued engagement. Potential Action Strategies: • Develop communications resources (e.g., sample messages, boilerplate text, images, and graphics) for REN members to share opportunities to encourage participation in the 30-day public comment period through social media, newsletters, websites, and other outlets. • Develop communications resources (e.g., sample messages, boilerplate text, images, and graphics) for REN members to share the final plan and to thank the community for participating in the process. • Solicit and document feedback from REN members on the effectiveness of the engagement of low-income residents and minority populations throughout the process. • Identify lessons learned for application to future transportation planning and decision-making processes. Resources • PolicyLink and USC Program for Environmental & Regional Equity. 2013. An Equity Profile of the Kansas City Region. http://www.marc.org/Regional-Planning/Creating-Sustainable-Places/assets/Kansas- City_Profile_23August2013.aspx. • MARC. Creating Sustainable Places: Equity Lens. http://www.marc.org/Regional-Planning/Creating- Sustainable-Places/assets/InitialEquityStrategiesforCorridors.aspx.

A-26 Equity Analysis in Regional Transportation Planning Processes Mid-Ohio Regional Planning Commission (MORPC) Agency Context The Mid-Ohio Regional Planning Commission (MORPC), which is the MPO for the Columbus, Ohio, region, has emerged as a national leader in transportation equity practice. The agency’s population-based approach to equity analysis, which uses travel-demand model outputs to analyze aggregate impacts of transportation actions on different demographic groups, provides a valuable alternative to the prevailing threshold-based approach. However, the agency’s equity practice has remained largely unchanged over the past decade. As MORPC looks ahead to its forthcoming 2020–2050 MTP cycle, agency staff are considering new and innovative ways to address the region’s equity issues. For example, Columbus has a relatively low unemployment rate, yet workplace access remains an issue for the region’s low-income populations and minority populations. In 2016, the City of Columbus won the U.S. DOT Smart City Challenge and was awarded $40 million to develop a “smart” transportation system that leverages technology and data to improve the movement of goods and people. As part of this initiative, the city is developing a multimodal trip planning application that provides a robust set of transit options and alternative travel options as well as routes, schedules, and dispatching availability. Use of this app will be supported by free public Wi-Fi access on buses and at transit stations. The city is also developing a common fare payment system that provides unbanked customers with access to bike and car sharing systems. To extend service for patrons who lack access to the trip planning app, Columbus will install multimodal trip-planning kiosks at select transit stations along the city’s recently launched bus rapid transit (BRT) corridor. Further, the city is developing an autonomous shuttle service in an area near the BRT corridor. (At some point in the future, this shuttle service could develop into a driverless ride-hailing program to better connect the city’s transit users to their destinations.) These projects are scheduled to launch in 2019 and 2020, and will be deployed first in neighborhoods with high concentrations of low-income populations or minority populations. The city’s overarching goal with these initiatives is to bridge first-mile/last-mile gaps and improve underserved persons’ lives through increased access to opportunity. Overview of Technical Assistance Columbus’s Smart City initiatives (http://smart.columbus.gov) will provide the city’s transit users with enhanced connectivity between travel modes, yet physical connectivity barriers surrounding these smart mobility hubs may hinder their full potential. Recognizing that safe and convenient interconnected bicycle and pedestrian networks are key to addressing first-mile/last- mile gaps, MORPC has been interested in testing promising peer agency techniques in multimodal network connectivity analysis. These types of analyses assess the characteristics and completeness of the bicycle, pedestrian, and highway networks in order to understand how they may impact travel behavior and route choice. The findings of these analyses can aid in identifying network connectivity gaps and understanding disparate levels of multimodal network connectivity between Pilot at a Glance Region: Columbus, Ohio MPO Service Population: 1,426,183 Phase: Long-Range Transportation Planning, Programming, Project Development Focus of Technical Assistance: Identify Needs and Concerns

Pilot Case Studies A-27 population groups and neighborhoods across a region. Validation and refinement of these findings through public engagement can be used alongside quantitative factors like residential density and job density to prioritize improvements to the multimodal network and improve regional equity outcomes. For the pilot, MORPC selected three points along the Cleveland Avenue BRT corridor that may become smart mobility hubs. These three points include the Linden Transit Center, the Northern Lights Transit Center and Columbus State Community College. These three points are either adjacent to low-income communities and minority communities along the BRT corridor or important travel destinations for these populations. The pilot assistance included analysis of walking potential within a half-mile of these three points; bicycling potential within 2 miles of these three points; and driving potential within 4 miles of these three points. The half-mile walk- shed distance and 2-mile bike-shed distance were chosen by MORPC under the assumption that transit patrons in these areas are willing to travel these distances on foot or by bike, respectively, to connect to a BRT station. (The distances also comport with travel-shed distances for these modes used in other studies on first-mile/last-mile accessibility.) The 4-mile drive-shed analysis served two purposes: 1. It demonstrated the number of residents or workers who could be served by a ride-hailing service or shuttle service within a given travel shed, and 2. It served as a proxy for the total multimodal network connectivity potential of the highway network, assuming all streets in a given travel shed included sidewalks and comfortable bicycling facilities. This provided the ratio of current bikeable or walkable streets to potential bikeable or walkable streets for a given travel shed. (For example, through this calculation one might find that 54% of streets within a half-mile driving distance from a given smart mobility hub include sidewalks). Method Broadly, this pilot sought to answer the following questions: 1. From an equity standpoint, what minority populations and low-income populations are currently served by the three mobility hubs, and how do the numbers compare? 2. What are the numbers of jobs accessible within each of these mobility hubs by a given mode? 3. Where are the gaps or opportunity areas in the current networks? 4. What minority populations and low-income populations could potentially be served by strategic bike and sidewalk infrastructure? In addition to the three mobility hubs along the BRT corridor, several different datasets were used to construct the transportation networks for this analysis: • The Sidewalk Inventory layer compiled by MORPC and available through the Open Data website (https://public-morpc.opendata.arcgis.com/datasets/) contains a comprehensive inventory of existing sidewalks and crosswalks. • The Bike Level of Comfort layer provided on MORPC’s Open Data website provides an indication of suitability for cyclists based on non-rush hour travel conditions and feedback from Columbus-area residents. • Open StreetMap (OSM) highway network data was obtained from Geofabrik (http://download.geofabrik.de/north-america/us/ohio.html), a firm specializing in OSM services. The road layer was used to supplement gaps or provide additional features (such as off-roadway paths) to the travel networks.

A-28 Equity Analysis in Regional Transportation Planning Processes • U.S. Census Bureau ACS 5-year data (2012–2016) at the block-group level was used to identify locations of households under the federal poverty level and current population counts of the minority population. • U.S. Census Bureau Decennial 2010 data at the block level was used to identify the minority population counts at a finer scale than is provided by block-group data from the ACS. • Longitudinal Employer-Household Dynamics (LEHD) Origin-Destination Employment Statistics (LODES) data Workplace Area Characteristics (WAC) from 2016 at the block- group level was used to identify the type of jobs available in a block. Using the Network Analyst extension with ArcGIS software, the three smart mobility hubs and three different mode networks were analyzed to create travel sheds using the corresponding travel distances for each mode. To calculate both the current and potential number of minority residents and low-income residents and jobs served by the various mode network travel sheds, these sheds were intersected with the census data described above. The demographic and job data were then apportioned based on the area within the various travel sheds. For example, if 50% of a given census block was enclosed within a half-mile walk shed, then half of the minority residents and low-income residents of that particular block would be considered to fall within the walk shed; this would be repeated for all other block groups that intersected the walk shed, and those demographic groups would be aggregated and compared to the control population within the same area. To identify the opportunity areas based on the potential surrounding a mobility hub, the number of low-income residents and minority residents that could be serviced by the road network—a proxy for the maximum possible potential of the multimodal network, excluding off-street paths— were compared to the bike sheds and walk sheds. Figure A-10 (next page) provides an example of the difference between an existing network and a potential network. The potential 2-mile bike shed represents a scenario wherein all existing roads are considered to be at least moderately comfortable for bicycle travel. Results Based on the findings of this analysis, both the Northern Lights and Linden Transit Centers have the potential to better connect over 500 minorities and over 300 low-income residents to the stations by expanding adjacent sidewalk infrastructure. Columbus State has the opportunity to connect to more jobs, but to much fewer residents. Likewise, the analysis finds that both the Northern Lights and Linden Transit Centers appear to have the opportunity to connect to neighborhoods with substantial minority populations and low-income populations with expanded bike infrastructure. Additionally, Linden has the most opportunity to connect Columbus residents to more jobs, with Columbus State close behind. Table A-3 depicts the potential bike-shed opportunity for low-income residents and minority residents and jobs for each of the three pilot study areas.

Pilot Case Studies A-29 Figure A-10. Current 2-mile bike shed compared to the potential 2-mile bike shed for the Linden Transit Center. Table A-3. Potential bike-shed opportunity by demographic groups for three mobility hubs. Mobility Hub Additional Minority Population Additional Low- Income Population Additional Total Population Additional Jobs Northern Lights 13,049 6,492 17,809 4,745 Linden 15,424 18,508 45,620 17,765 Columbus State 3,621 2,607 9,987 14,064 Lessons Learned: Implications for the Equity Guide Network data should be compared to on-the-ground conditions before making any conclusions. Key connections that currently exist may be missed for various reasons in the network layer. For example, if a half-mile walk shed and 2-mile bike shed terminate at the same point, this may represent a barrier such as a river or train tracks. However, this could also represent an inaccuracy in the network or a short break in pedestrian or bike network that would not represent a considerable barrier in reality (for the purpose of this analysis MORPC opted to take a conservative approach to defining the bicycle level of comfort network). Engaging affected populations to learn about experience in the field is a key step to ground-truthing and refining these findings. Resources • City of Columbus, Ohio. Smart Columbus. https://smart.columbus.gov/ • FHWA. 2018a. Guidebook for Measuring Multimodal Network Connectivity. https://www.fhwa.dot. gov/environment/bicycle_pedestrian/publications/multimodal_connectivity/. • FHWA. 2018b. Guidebook for Measuring Multimodal Network Connectivity, Appendix: Case Studies. https://www.fhwa.dot.gov/environment/bicycle_pedestrian/publications/multimodal_connectivity/appendix.cfm.

A-30 Equity Analysis in Regional Transportation Planning Processes Technical Addendum to MORPC Pilot Study This addendum provides a detailed discussion of the methods, results, level of effort, and limitations of the MORPC multimodal network connectivity pilot case study featured in the guidebook. Analysis Context Columbus’s forthcoming Smart City initiatives (http://smart.columbus.gov) seek to provide the city’s transit users with enhanced connectivity between travel modes. Yet physical connectivity barriers surrounding these smart mobility hubs may hinder their full potential. Multimodal network connectivity analyses assess the characteristics and completeness of the bicycle, pedestrian, and highway networks in order to understand how these factors may impact travel behavior and route choice. The findings of these analyses can aid in identifying network connectivity gaps and understanding disparate levels of multimodal network connectivity between population groups and neighborhoods across a region. For the pilot, MORPC selected three points along the Cleveland Avenue BRT corridor that may become smart mobility hubs (Figure A-11). These three points include the Linden Transit Center, the Northern Lights Transit Center and Columbus State Community College. The points are either adjacent to underserved communities (neighborhoods with higher percentages of minority individuals and/or individuals with low incomes) along the BRT corridor or important travel destinations for these populations. The pilot assistance included analysis of walking potential within a half-mile of these three points; bicycling potential within 2 miles of these three points; and driving potential within 4 miles of these three points. Analysis Process Data Assembly Several different datasets were used to construct the transportation networks for this analysis: • MORPC staff provided the locations of three mobility hubs along the BRT corridor. • The Sidewalk Inventory layer compiled by MORPC and available through the Open Data website contains a comprehensive inventory of existing sidewalks and crosswalks. • The Bike Level of Comfort layer provided on the Open Data website provides an indication of suitability for bicyclists based on non-rush hour travel conditions and feedback from Columbus-area residents. • OSM data is available through a variety of sources. For this pilot study, the road layer for Ohio was obtained from Geofabrik, a firm specializing in OSM services. The road layer was used to supplement gaps or provide additional features (such as off-roadway paths) to the travel networks. Technical Analysis at a Glance Focus: Multimodal First-Mile/Last- Mile Connectivity Analysis Context of Study Area: Urban Scale of Study Area: Three Pilot Mobility Hubs Tools & Models: GIS Application: Long-Range Planning, Programming, Project Development Equity Analysis Step: Identify Needs and Concerns

Pilot Case Studies A-31 • U.S. Census Bureau ACS 5-year data (2012–2016) at the block-group level was used to identify locations of households under the federal poverty level and current population counts of the minority population. • U.S. Census Bureau Decennial 2010 data at the block level was used to identify the minority population counts at a finer scale than is provided by block-group data from the ACS. • LEHD Origin-Destination Employment Statistics (LODES) data WAC from 2016 at the block level was used to identify the types of jobs available in a block. • OSM road data, the sidewalk inventory, and the bike level of comfort inventory were used to compile networks for each of the three modes: walking, biking, and driving. Each mode required a different set of considerations. Figure A-11. MORPC study area, underserved communities and mobility hubs.

A-32 Equity Analysis in Regional Transportation Planning Processes Table A-4. MORPC pilot study GIS layers. Mode Layers Used Feature Types Walk • MORPC Sidewalk inventory • OSM roads • All sidewalk types excluding “X” or no sidewalks • OSM fclass types of footway, path, pedestrian, steps, and service Bike • MORPC bike level of stress • OSM roads • Good, Moderate, and Residential types • OSM fclass types of residential, cycleway, and path Drive (Ride Hail) • OSM roads • All roads excluding motorway and motorway_link, path, pedestrian, track, track_grade1, track_grad2, steps, footway, and cycleway • Walking: The pedestrian network was compiled using the sidewalk inventory excluding locations where no sidewalks are present, and the OSM road layer (see Table A-4). The sidewalk inventory was supplemented with OSM data to provide additional off-roadway features. The rationale for including these features is that they provide key links through parks, campuses, and other areas. These features were relevant to include in this particular study area; however, when moving to other study areas the data will need to be reviewed to ensure the selection aligns with ground conditions. • Biking: The biking network was compiled using the bike level of comfort and the OSM road layers. The level of comfort rating was completed on major roadways and contains ratings of residential, good, moderate, and poor. Residential, good, and moderate roadways were selected for the analysis, and supplemented by OSM residential road data in order to capture all potential comfortable bike routes. • Driving: The ride-hailing network was compiled using the OSM road data and excluded pedestrian features and motorways. The rationale for excluding motorways is that this study aims to assess first- and last-mile transit connections and not travel from a substantial distance away on grade-separated highways. Analyzing Travel Sheds Using the Network Analyst extension with ArcGIS software, the mobility hubs and three different mode networks were analyzed to create travel sheds for the various travel distances selected by MORPC. These travel sheds can be viewed in Figure A-12 through Figure A-15. • Walking: A half-mile walk shed (about a 10-minute walk assuming an average walking speed of 3.1 mph), • Biking: A 2-mile bikeshare travel shed, and • Driving (ride-hailing): A 4-mile drive shed. To create the travel sheds, connected networks were buffered by 100 meters—the agreed-on distance for comfortably accessing a network route. This does present a limitation to the analysis, because often paths for accessing locations off the road network (such as through an apartment building courtyard) are not adequately mapped in inventory datasets, resulting in missed populations during the overlay analysis.

Pilot Case Studies A-33 Figure A-12. Northern Lights Transit Center current drive, bike, and walk sheds.

A-34 Equity Analysis in Regional Transportation Planning Processes Figure A-13. Linden Transit Center current drive, bike, and walk sheds.

Pilot Case Studies A-35 Figure A-14. Columbus State Community College Transit Center current drive, bike, and walk sheds.

A-36 Equity Analysis in Regional Transportation Planning Processes Purpose The generated travel sheds are used to answer two key questions in this study: 1. From an equity standpoint, what percentage of residents currently served by the three mobility hubs are considered minority or low-income individuals, and how do the numbers compare to non-minority/mid- or high-income? 2. Where are the gaps or opportunity areas in the current networks? What minority and low- income populations in the study area could be served by strategic bike and sidewalk infrastructure? In addition to underserved communities, the estimated number of jobs within the travel sheds were calculated using the LODES data to assess the demand for people to travel to the mobility hub, rather than from the hub. This metric is important to consider alongside residential characteristics. To answer the second question and identify the opportunity areas based on the potential surrounding a mobility hub, two additional travel sheds were created to represent the potential walking and biking network. These were created from the road network rather than using a straight distance measure. The road network was seen as a good indication of all potential sidewalk and bike lanes that could exist on the existing network. It is worth noting that the bike and pedestrian network may potentially “beat” the road network in certain areas if there is a path unavailable for car use, so including these existing features in the potential network should be considered. An example of the difference between an existing network and a potential network can be seen in Figure A-15, which shows the current 2-mile bike shed and a potential 2-mile bike shed for the Linden Transit Center. In Figure A-15, the potential bike shed represents conditions wherein all existing roads are considered comfortable for bike travel. To calculate both the current and potential number of minority and low-income residents (and also jobs) within the travel sheds, the various travel sheds were intersected with the best available census data. This varied depending on the metric. Minority and low-income populations were assessed at the block-group level. To further refine the distribution of the populations within the block groups, census blocks with zero population were erased from the block group data. Jobs data was available at the block level. The block-group and block totals were apportioned based on the area within the various travel sheds. Figure A-15. Current 2-mile bike shed compared to the 2-mile bike shed for the Linden Transit Center.

Pilot Case Studies A-37 Results: Mobility Hub Comparison The results of the analysis aimed to compare the three hubs and answer the first of the two key questions of this analysis: How many residents currently served by the existing networks are considered minority or low-income? Minority Communities • Walking: Of the three study hubs, the Linden Transit Center’s current walk shed serves the highest minority population, total population, and minority percentage. The Columbus State Transit Center’s current walk shed, with 279 total minority residents, serves the lowest population in all three categories (Table A-5). Table A-5. Walk sheds—minority populations. Mobility Hub Minority Population Total Population Percentage Minority Northern Lights 529 1,161 46% Linden 1,223 1,734 71% Columbus State 279 824 34% • Biking: The Columbus State Community College Transit Center bike shed has the lowest minority percentage but serves the highest number of minority people, whereas the Linden Transit Center serves the least number of overall people but the highest minority percentage (Table A-6). This may indicate a potential opportunity to better serve minority communities, which will be explored in the next section. Table A-6. Bike sheds—minority populations. Mobility Hub Minority Population Total Population Percentage Minority Northern Lights 9,394 21,410 44% Linden 5,466 6,637 82% Columbus State 11,733 30,315 39% • Driving (ride-hailing): The Northern Lights Transit Hub serves the highest minority population within a 4-mile ride-hailing shed, both in total people and percentage (Table A-7). Linden and Columbus State serve a higher total population overall. Table A-7. Drive (ride-hailing) sheds—minority populations. Mobility Hub Minority Population Total Population Percentage Minority Northern Lights 77,066 150,320 51% Linden 60,056 165,618 36% Columbus State 60,227 170,467 35%

A-38 Equity Analysis in Regional Transportation Planning Processes Low-Income Communities • Walking: Of the three study hubs, the Linden Transit Center current walk shed serves the highest low-income population, total population, and low-income percentage (Table A-8). Table A-8. Walk sheds—low-income populations. Mobility Hub Low-income Population Total Population Percentage Low-income Northern Lights 478 1,161 41% Linden 1,040 1,734 60% Columbus State 254 824 31% • Biking: The Columbus State Community College Transit Center bike shed serves the highest number of low-income and total population, whereas the Linden Transit Center serves the least number of overall people but the highest low-income percentage (Table A-9). Table A-9. Bike sheds—low-income populations. Mobility Hub Low-income Population Total Population Percentage Low-income Northern Lights 5,525 21,410 26% Linden 3,173 6,637 48% Columbus State 8,389 30,315 28% • Driving (ride-hailing): The sheds show modest differences across population totals, with Northern Lights serving fewer low-income residents (Table A-10). Table A-10. Drive (ride-hailing) sheds—low-income populations. Mobility Hub Low-income Population Total Population Percentage Low-income Northern Lights 37,459 150,320 25% Linden 47,512 165,618 29% Columbus State 52,044 170,467 31% Jobs Jobs are an important indicator of what might draw people to a particular mobility hub and complete a “last-mile” trip, rather than residents completing a first-mile trip. Stations such as Columbus State Community College provide access to more employment opportunities for residents of the broader region than to residents nearby. Table A-11 compares the numbers of jobs within proximity to each hub, which varies depending on the mode shed.

Pilot Case Studies A-39 Table A-11. Jobs within travel sheds of stations. Mobility Hub Walk Shed Bike Shed Ride Shed Northern Lights 697 2,749 60,792 Linden 841 1,294 188,494 Columbus State 10,339 109,170 200,380 Results: Potential Opportunities Table A-12 and Table A-13 aim to answer the second key question of this analysis by providing the totals for additional residents who could be served. These are the numbers of individuals living within the optimal potential sheds for the respective mode distances. Higher numbers indicate opportunity areas for connecting more people to the transit hub with expanded walking or biking infrastructure. It is important to note that network data should be compared to on-the-ground conditions before making any conclusions. Key connections that exist currently may be missed due to various reasons in the network layer, which will be discussed in the section titled “Limitations.” • Walking: Both the Northern Lights and Linden Transit Centers appear to have the opportunity to better connect over 500 minorities and over 300 low-income residents to the stations with expanded sidewalk infrastructure, with Northern Lights connecting a larger total population. Columbus State has the opportunity to connect to more jobs, but much fewer residents. Figure A-16 depicts both the current and potential walk sheds for Northern Lights overlaid on block-group minority populations. Sidewalk gaps along many of the streets prevent the shed from extending to the northeast. As seen in Figure A-16, block groups in all directions may potentially benefit from further completion of sidewalks. Table A-12. Potential walk-shed opportunities. Mobility Hub Additional Minority Population Additional Low- income Population Additional Total Population Additional Jobs Northern Lights 586 415 1,269 429 Linden 578 307 807 557 Columbus State 136 119 407 3,526 • Biking: Both the Northern Lights and Linden Transit Centers appear to have the opportunity to connect a large minority population with expanded bike infrastructure, with Linden adding a substantial increase to both the low-income population (by more than 18,000 people) and total population (by more than 45,000 people). Additionally, Linden has the most opportunity to connect workers with more jobs, with Columbus State close behind. Figure A-17 depicts both the current and potential bike sheds for Linden Transit Center overlaid on block-group minority populations. Poor level of comfort streets traveling north, east, and south of the station appear to prevent a large number of people from having comfortable bike access to the station. Block groups to the west and south of the existing bike network contain a high percentage of minorities; however, as seen in Figure A-17, these groups appear to have a tough time connecting to the current bike network.

A-40 Equity Analysis in Regional Transportation Planning Processes Table A-13. Potential bike-shed opportunities. Mobility Hub Additional Minority Population Additional Low- income Population Additional Total Population Additional Jobs Northern Lights 13,049 6,492 17,809 4,745 Linden 15,424 18,508 45,620 17,765 Columbus State 3,621 2,607 9,987 14,064 Figure A-17. Potential opportunity areas for greater connection of minority communities to the Linden Transit Center bike network. Figure A-16. Potential opportunity areas for greater connection of minority communities to the Northern Lights Transit Center walking network.

Pilot Case Studies A-41 Methodology The steps below outline the specific steps to perform this analysis. One of the key aims of this study was to derive a methodology that is both repeatable and scalable across many mobility hubs, or different focal locations (e.g., hospitals). All existing network data (OSM, sidewalk inventory, bike level of comfort) and mobility hub locations were loaded into a single GIS map. ESRI ArcGIS software was selected for this analysis because of its Network Analyst extension and its familiarity to GIS analysts at most agencies. The travel networks for walking, biking, and driving were prepared using the following steps: 1. Clip to the study area, defined as a 5-mile buffer around the mobility hubs for this analysis. This distance encompasses the maximum of 4 miles for the drive shed, while still depicting what is present just beyond the study distances. 2. Select the appropriate features using queries. 3. Merge the data obtained from MORPC’s open data portal with supplemental OSM data. The MORPC inventories were completed for a certain purpose (such as identifying a level of stress on major streets only), making it necessary to supplement the data to build a more complete network. 4. The Feature to Line tool was then used to “fix” any topography errors resulting from the merge, such as lines not being split when crossed by another line. If this step is not taken, when creating the travel shed turns will not be modeled appropriately. This tool will automatically add nodes at all intersecting lines. 5. The Split Lines at Vertices tool should then be used to ensure that nodes are placed where a line touches but does not cross another line. The Feature to Line tool will take care of four- way intersections, whereas the Split Lines at Vertices tool will take care of three-way intersections. After the line layers for the three modes were built following the previous steps, a network dataset can be built and “solved” using the Network Analyst extension. This is the process taken to create the various travel sheds. 1. Create a new geodatabase, and then add a new feature dataset for each of the different modes being analyzed. Separate feature datasets are required because only one network dataset is allowed within a single feature dataset. Take note of the unit of measurement with the selected coordinate system (use feet or meters). 2. Add each of the base network layers described in the Preparing Data section to its respective feature dataset created in Step 1. 3. Right click on the Feature Dataset name when in the Catalog and select New >> Network Dataset. Use Length as the cost, allow for global turns, and build a regional index. One-way streets were not modeled in the analysis. 4. It is important to note that if changes are made to the network at a later time, the network dataset will need to be recreated. The Build Network tool can be used to reconstruct an existing network. The following steps outline how to create the travel-shed polygon layers using the network datasets created in Step 3. It is highly recommended that the following steps be done in the ArcGIS ModelBuilder environment so that all selected parameters and input datasets can be documented. Additionally, travel sheds are rarely perfect the first time, and viewing them can draw attention to obvious flaws in the underlying network. Using the ModelBuilder environment makes for quick

A-42 Equity Analysis in Regional Transportation Planning Processes repeatability of steps after the input dataset has been fixed. Once built, the model can also be copied for analyzing additional mode networks. 1. Use the Make Service Area Layer tool with the network dataset created in Step 3 as the input. There are other ways of building travel sheds, however, this was the approach selected for this particular analysis. 2. A number of options need to be set within the tool. The first is the Impedance Attribute, which should be set to Length. Travel From or To Facility is important when modeling one- way features and looking into differentiating between residential characteristics (travel to) or opportunity and job characteristics (travel from), however, for this particular analysis this setting was not used. Only one shed was generated per mode not modeling one-way streets. 3. Default Break Values allow the assessment of multiple travel distances at one time. Ensure that the units selected are the same as the feature dataset coordinate system. 4. For this analysis, lines were selected for the travel shed instead of polygons and then buffered by 100 m to create consistent polygons. This was the distance determined to be reasonable to walk or bike to get on a comfortable network. 5. The next step is to Add Locations to the service area. These are the mobility hub centers that are being assessed. A search tolerance may need to be set so that the point is able to “get on” the network. Note that during repeat runs of the model, this step would often throw an error. To get past this, simply remove and re-add the step to the model. 6. Next, use the Solve tool to create a temporary solution in memory to the service area based on the input locations. The Select Data tool can then capture individual components of the solution, such as lines and polygons. This selection is only in memory for the time being. A final use of the Select tool can save the output for use in the overlay analysis with the census areas. After the generated travel sheds have been reviewed for accuracy and determined to be sufficient, these polygons can be overlaid with population or employment data to estimate who is being serviced by the current network. The steps below provide a general outline of how this was done. 1. Census and LODES data was joined with block group or blocks depending on the resolution available. Some data that is sensitive (e.g., income) is not available at the block level; however, general population estimates are. 2. Blocks determined to have zero population were erased from the block-group level data. This helped to further refine the distribution of populations at the block-group level by eliminating certain areas (e.g., transportation corridors, parks, industrial areas). 3. The block groups were then intersected with the travel sheds to obtain minority and low- income population estimates, and blocks intersected to provide number of jobs estimates. 4. The block-group totals were apportioned based on the area within the travel shed. For example, if the block group is 50% within a shed, only 50% of the population was counted. Level of Effort Table A-14 outlines high-level steps, basic assumptions, estimated hours (by mode, if applicable), and level of skill required. This analysis does require experience working with Network Analyst, complex network datasets, and census data. If the analyst does not have this experience, more time should be expected to complete a similar study. More time also should be added if studying many locations at a time for reviewing data and compiling results.

Pilot Case Studies A-43 Table A-14. Level of effort for MORPC GIS analysis. Step Assumptions Estimated Hours Skill Required Compile multiple network datasets into a single layer to depict the network available by mode • Data are readily available and do not require editing • Network criteria are pre- established ~5 hours per mode; more to establish and quality control criteria Low (moderate if criteria are not pre- established) Acquire and set up census and LEHD data • Data are not readily available and need to be obtained • Analyst has general familiarity with census data fields, equity characteristics, and joining tabular data to spatial units ~10 hours Moderate Build network datasets using Network Analyst and solving them to define routes and travel sheds • Access to Network Analyst extension • Familiarity with network dataset characteristics and limitations • Experience with ArcGIS ModelBuilder environment ~10 hours per mode Moderate to advanced Check outputs, identify and resolve issues, and rerun analysis steps Steps have been built in a model environment ~5 hours per mode Moderate to advanced Overlay analysis with residential and employment data, apportion and compile data to shed ~5 hours per mode Moderate to advanced Limitations This study was based off of existing network data and did not set out to “fix” any network issues that may have produced erroneous results. This was intentional for a few reasons. The first goal was to assess how readily usable existing data can be to performing this type of analysis and the second goal was to highlight potential limitations of the data that may help guide how future inventories are compiled.

A-44 Equity Analysis in Regional Transportation Planning Processes A few of the main limitations identified in this analysis are as follows: • Data from different sources may not always align. For routing analyses to work correctly, data needs to be connected; otherwise, the break will be a barrier to further travel. Snapping datasets together may be one potential fix to this issue. • Often network datasets do not accurately capture all available routes to all modes, with preference often given to roads for car use. Obvious paths for pedestrians and bicyclists may be missing, such as connections to apartment buildings, around campuses, or through parks. This can greatly reduce the walk and bike sheds compared to on-the-ground conditions. Sidewalk and bike level of comfort inventories should aim to contain all available routes, not just those along existing road networks. • The 100 m distance from routes selected for this analysis may not accurately depict the distance that people are willing to travel to get on a network and may drastically undercut populations that are disconnected due to the previous limitation. For example, if a path traveling to a group of apartments is not in the network, this block group may be erroneously severed from the shed. • Data at the block-group level could be further refined to provide a more accurate distribution of where individuals from minority and low-income populations reside. This could be done by using parcel level data rather than distributing individuals evenly across block groups. • Road crossings are not accurately depicted. In Figure A-18, Cleveland Avenue is identified as having a poor level of comfort rating for bicyclists. The neighboring residential streets, however, are likely quite comfortable. Because of how the network data is set up, the line is not traveling through the intersection, and bicyclists will not be able to cross Cleveland Avenue in the routing analysis. One potential solution to this issue is to supplement the bike network with crosswalk data, which may help bridge some of these gaps. This does, however, require the crosswalks to be attributed separately from sidewalks so that they are easy to extract from a sidewalk inventory. Figure A-18. Inaccurate depiction of road crossing resulting in a poor level-of-comfort rating for bicyclists.

Abbreviations and acronyms used without definitions in TRB publications: A4A Airlines for America AAAE American Association of Airport Executives AASHO American Association of State Highway Officials AASHTO American Association of State Highway and Transportation Officials ACI–NA Airports Council International–North America ACRP Airport Cooperative Research Program ADA Americans with Disabilities Act APTA American Public Transportation Association ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASTM American Society for Testing and Materials ATA American Trucking Associations CTAA Community Transportation Association of America CTBSSP Commercial Truck and Bus Safety Synthesis Program DHS Department of Homeland Security DOE Department of Energy EPA Environmental Protection Agency FAA Federal Aviation Administration FAST Fixing America’s Surface Transportation Act (2015) FHWA Federal Highway Administration FMCSA Federal Motor Carrier Safety Administration FRA Federal Railroad Administration FTA Federal Transit Administration HMCRP Hazardous Materials Cooperative Research Program IEEE Institute of Electrical and Electronics Engineers ISTEA Intermodal Surface Transportation Efficiency Act of 1991 ITE Institute of Transportation Engineers MAP-21 Moving Ahead for Progress in the 21st Century Act (2012) NASA National Aeronautics and Space Administration NASAO National Association of State Aviation Officials NCFRP National Cooperative Freight Research Program NCHRP National Cooperative Highway Research Program NHTSA National Highway Traffic Safety Administration NTSB National Transportation Safety Board PHMSA Pipeline and Hazardous Materials Safety Administration RITA Research and Innovative Technology Administration SAE Society of Automotive Engineers SAFETEA-LU Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005) TCRP Transit Cooperative Research Program TDC Transit Development Corporation TEA-21 Transportation Equity Act for the 21st Century (1998) TRB Transportation Research Board TSA Transportation Security Administration U.S. DOT United States Department of Transportation

TRA N SPO RTATIO N RESEA RCH BO A RD 500 Fifth Street, N W W ashington, D C 20001 A D D RESS SERV ICE REQ U ESTED ISBN 978-0-309-48167-0 9 7 8 0 3 0 9 4 8 1 6 7 0 9 0 0 0 0 Equity A nalysis in Regional Transportation Planning Processes, Volum e 1: G uide TCRP Research Report 214 TRB