An Assessment of Data, Tools, and Metrics for Equity in Decisions About Surface Transportation Investments (2025)
Chapter: 4 Emerging Practices and Innovations That Hold Promise
4
Emerging Practices and Innovations That Hold Promise
In reviewing the state of practice, the committee identified emerging practices or innovations related to data, metrics, or analytic methods, many of which have the potential to advance equity assessment of surface transportation projects and programs. This chapter reviews these promising practices and innovations, with special attention to their ability to improve the evaluation of a project or program effectiveness for underserved or historically disadvantaged communities. The chapter begins by providing an overview of the indicators for proximity to destinations, sometimes called opportunity indicators, that can be used to assess disparities in access to employment, health care, education, and other destination types. Next, the chapter introduces the concept of “transportation insecurity”—the inability to travel to meet daily needs safely and reliably—and a specific method for its measurement, followed by a discussion of its potential uses for project and program decision making.
For health and other environmental justice impacts, the committee examined emerging practices and innovations in three areas: housing and transportation cost burdens, transportation safety and security, and air quality. For housing and transportation cost burdens, which also affect people’s access to housing and access to destinations, the chapter reviews methods that assess these cost burdens both together and separately. The chapter then discusses the potential for these cost burden indicators to assist government in working across the transportation and housing sectors, for the benefit of historically disadvantaged or underserved communities. The committee discussed indicators and data for transportation safety and security in the context of safe system approaches, while also drawing on
practices from public health. Air quality is presented as an example of an issue where new, lower-cost methods of data collection and analysis can be used to expand understanding of disproportionate impacts as well as direct solutions.
PROXIMITY TO DESTINATION INDICATORS
Two indicators for proximity—in time as well as space—to destinations are well established in the research literature but have lagged in practice. These kinds of metrics are now gaining wider use in practice and hold promise for assessments related to equity and the statement of task’s interest in indicators “for safe and reliable access to housing, employment, health care, education, and essential services.” Partially as a result of the plethora of data available and enhanced computing power, it is now much more feasible to integrate these metrics into transportation decision making.1 As noted in Chapter 2, causal chain analysis underpinned much of the committee’s thinking about the effects of transportation investments on societal outcomes. The committee acknowledges that it is difficult to link a change in a proximity metric directly to a change in societal outcome. However, understanding the relationship between, say, improved bus access and, for example, better access to educational opportunities, which can ultimately improve educational attainment, is critical for selecting the best metric for the context being evaluated.
Cumulative opportunity indicators count the number of destination types (e.g., jobs or education) within a travel time threshold from a specific location via a specific mode (e.g., number of jobs within 45 minutes of a census block group via driving or transit). Cumulative opportunity indicators can also be flipped to determine the number and location of residents within a given travel threshold of a destination, such as the number or percent of residents within a 20-minute walk of a public library and the locations of residents outside this threshold. Cumulative opportunity indicators typically use travel time thresholds. This travel time threshold can be a simple boundary that counts all destinations within the threshold. However, many cumulative opportunity analytic tools incorporate the travel time threshold into a decay function that values destinations nearer to the
___________________
1 For an assessment of accessibility measures in general, refer to Siddiq, F., and B. D. Taylor. 2021. “Tools of the Trade?: Assessing the Progress of Accessibility Measures for Planning Practice.” Journal of the American Planning Association 87(4):497–511. https://doi.org/10.1080/01944363.2021.1899036. Refer also to National Academies of Sciences, Engineering, and Medicine. 2022. Accessibility Measures in Practice: A Guide for Transportation Agencies. Washington, DC: The National Academies Press. https://doi.org/10.17226/26793.
origin more highly than destinations farther away, until with distance the value of a destination “decays” to zero.2
Travel time opportunity indicators produce the travel time to a specified location via a range of modes, such as the travel time from a census block to the nearest hospital or third nearest grocery store via driving or transit.3 Travel time opportunity measures can produce a profile of residential service levels, such as the percent of residents who can reach three grocery stores within 20 minutes and the locations of residents who cannot or the residential locations that are twice the average travel time to a hospital.4
Finally, these proximity indicators can be thought of as “opportunity” indicators in part because they are based on spatial analysis to potential destinations. On their own, they do not account for other factors that may determine whether someone travels to a destination, such as inadequate infrastructure that prevents someone with a disability from walking or rolling, or fear while traveling of crime, sexual harassment, or unsafe roads. Trip cost is another factor that may prevent someone from accessing an opportunity. Although opportunity indicators could use trip cost instead of or in combination with travel time, most opportunity indicators rely on travel time.
Uses for Equity Analysis
Early development of opportunity indicators for practice focused on access to jobs and used cumulative opportunity indicators. Although jobs are still the most frequently analyzed destination type, new data sources allow analysis of a significantly expanded set of destinations such as grocery stores, health care facilities, education sites, parks, community centers, and social service providers. When these opportunity indicators are analyzed in conjunction with demographic and mode information, opportunity measures can be used to assess equity in access.5
___________________
2 For a more detailed explanation of decay functions and their uses, see Conveyal. “Decay Functions.” https://docs.conveyal.com/learn-more/decay-functions. Accessed July 15, 2024.
3 Travel time opportunity measures are also called dual access measures. Cui, M., and D. Levinson. July 2020. “Primal and Dual Access.” Geographical Analysis 52(3):452–474. https://doi.org/10.1111/gean.12220.
4 For a study that uses both cumulative opportunity and travel time opportunity measures see Accessibility Observatory, University of Minnesota. May 2024. “Access in Appalachia: Pilot Implementation Project.” https://www.cts.umn.edu/publications/report/access-in-appalachia-pilot-implementation-project.
5 For an example of how to use cumulative and travel time opportunity measures and analysis methods to conduct population-based disparity analysis, see Accessibility Observatory, University of Minnesota. May 2024. “Access in Appalachia: Pilot Implementation Project.” Pp. 25–26. https://www.cts.umn.edu/publications/report/access-in-appalachia-pilot-implementation-project.
An emerging practice is to flip opportunity indicators and use them to identify the locations of populations of concern. In the Transportation Equity Index developed by California Department of Transportation (Caltrans), “Access to Destinations” is one of three “screens” that make up the index. The Access to Destinations screen consists of four cumulative opportunity metrics mapped on the census block level, where higher ratios indicate blocks with relatively higher multimodal access to destinations.6
- Transit Access to Destinations—Work: the ratio of transit access to jobs to congested auto access to jobs, focused on regional-scale trip making.
- Transit Access to Destinations—Non-Work: the ratio of transit access to nonwork destinations to auto access to nonwork destinations, focused on regional-scale trip making.
- Bicycle Access to Non-Work Destinations: the ratio of bicycle access to nonwork destinations on the low-stress street network to bicycle access to nonwork destinations on the high-stress street network, focused on local trip making.
- Pedestrian Access to Non-Work Destinations: the ratio of pedestrian access to nonwork destinations on the street network to pedestrian access to nonwork destinations on the “as-the-crow-flies” network, focused on local trip making.
USDOT’s Equitable Transportation Community (ETC) Explorer’s Transportation Access Index uses a similar approach. The index has six indicators, three of which are calculated using opportunity metrics: the jobs within a 45-minute drive and the estimated average drive time and walk time to points of interest; these are all measured at the census block group level. For the points of interest analysis, the ETC Explorer uses the average drive or walk time to the two nearest grocery stores, parks, medical facilities, and adult education sites.7 Similarly, Transit Center’s Equity Dashboard offers cumulative opportunity metrics, travel time opportunity metrics, and the ratio of transit access to auto access. Available for select metropolitan
___________________
6 The Caltrans EQI uses travel-time weighted cumulative opportunity measures with a decay curve of 30 minutes for the transit indicators and 15 minutes for the bicycle and pedestrian indicators. Caltrans. March 2024. “Caltrans Transportation Equity Index (EQI): Version 1.0 Documentation.” https://dot.ca.gov/-/media/dot-media/programs/esta/documents/race-equity/eqi/v1/030124eqidocumentationv10a11y.pdf.
7 U.S. Department of Transportation (USDOT). 2024. “ETC Explore Indicators.” Updated December 4. https://www.transportation.gov/priorities/equity/justice40/etc-explorer-indicator-table. See also USDOT. 2023. “Equitable Transportation Community (ETC) Explorer: ETC Explorer Technical Documentation.” May. https://www.transportation.gov/sites/dot.gov/files/2023-05/5.2.23ETC%20Explorer%20Technical%20DocumentationFinal.pdf.
regions as a demonstration project, the dashboard offers analysis of transit access to jobs, park space, hospitals, urgent care centers, pharmacies, colleges and universities, supermarkets, and early voting centers. Its transit access analysis includes a fare-constrained option examining only at trips costing less than $4–$5, depending on the metropolitan region.8
Opportunity measures have also been incorporated into project selection processes. Virginia’s Smart Scale project selection process, in place since 2016, includes the proposed project’s change in proximity to destinations as components of its “accessibility” factor, which itself is one of six factors for the project selection evaluation system. In addition to an access-to-transportation measure, this accessibility factor includes cumulative opportunity measures for access to jobs for total population and access to jobs for disadvantaged populations within 45 minutes by driving or 60 minutes by transit, walking, or bicycling. Cumulative access to jobs is measured with a decay function. Driving to jobs is measured at the block-group level, and taking transit, walking, or bicycling to jobs is measured at the block level.9
Opportunity measures can inform long-range transportation plans as well as project selection. The Wasatch Front Regional Council (WFRC), the metropolitan planning organization (MPO) for the Salt Lake City region, has been using opportunity indicators as part of its long-range transportation planning and improvement programs since 2017. These metrics are also an important component of WFRC’s equity planning work.10 At the regional scale, WFRC uses its travel demand model and cumulative opportunity metrics to analyze access to jobs and access to workers. At the local scale, WFRC has developed a geographic information system of the multimodal local transportation network and the locations of schools, groceries, community centers, health care and social service providers, and parks. Using these tools, the opportunity metrics inform scenario analysis for the regional vision and the development and prioritization of regional-scale projects as well as smaller projects affecting subdistricts or neighborhoods.11
Several state DOTs and the Washington, DC, District Department of Transportation are investing in their capacity to conduct access to destination analysis as partners in the National Accessibility Evaluation “pooled fund” study that began with cumulative opportunity analysis of jobs and
___________________
8 TransitCenter. 2021. “About the TransitCenter Equity Dashboard.” https://dashboard.transitcenter.org/about.
9 Commonwealth Transportation Board. 2024. “SMART SCALE Technical Guide.” Revised February. https://smartscale.virginia.gov/media/smartscale/documents/508_R6_Technical-Guide_FINAL_FINAL_acc043024_PM.pdf.
10 Wasatch Front Regional Council (WFRC). “Equity Planning.” https://wfrc.org/public-involvement/equity-planning. Accessed July 15, 2024.
11 WFRC. “Access to Opportunities.” https://wfrc.org/maps-data/access-to-opportunities. Accessed July 15, 2024.
has now expanded to include grocery stores, childcare, educational destinations, health care, leisure opportunities, and freight nodes.12 Produced by the University of Minnesota Accessibility Observatory, the data are designed to be used for both cumulative opportunity and travel time opportunity analyses and provide nationwide coverage at the census-block level. The study, funded through 2025 with plans to continue longer, delivers annual statewide access to destinations data at the block level using updated information on destination locations, streets networks, and travel times for automobile, transit, walking, and biking. State-specific data are made available to pooled fund partners to download for analysis and visualization. A custom online mapping tool, in iterative development, was delivered to all study partners in 2024, allowing users to visualize access to essential destinations at the census-tract level by auto, bike, transit, and walk modes.13 The availability of annual updates allows states to track changes over time.
Data, Analysis, and Capacity Needs
USDOT’s ETC Explorer, University of Minnesota’s Accessibility Observatory, and other commercial providers have begun the work of making these proximity-to-destinations indicators more widely available for an expanding set of destination types. Travel demand models can build on these measures, for example, by calculating proximity via other modes or by forecasting the impacts of transportation system improvements and other interventions. However, limitations in commonly used travel demand model applications and staff technical capacity present a barrier to increasing the use of opportunity metrics at the metropolitan level. In addition, travel demand models are more commonly used at the MPO level and far less by state departments of transportation (DOTs).
The National Accessibility Evaluation pooled fund study described earlier provides census block–level data on multimodal access to essential destinations for existing conditions. However, access is limited to participating states. Forecasting a transportation project’s or program’s impacts on access could be done with a travel demand model in urban areas or with a
___________________
12 The 11 participating states are California, Connecticut, Florida, Illinois, Maryland, Massachusetts, Michigan, Minnesota, North Carolina, Texas, and Virginia; the University of Minnesota’s Accessibility Observatory provides the data and analysis capabilities. “National Accessibility Evaluation Phase II Access Across America.” Transportation Pooled Fund, updated February 1, 2022. https://pooledfund.org/Details/Study/682. Accessed November 23, 2024.
13 Accessibility Observatory, University of Minnesota. “Minnesota Interactive App (NAE 2022): Access to Opportunities Using Census Population Centers at Tract Level.” https://umn.maps.arcgis.com/apps/dashboards/2f66bd8906ed47de9c9d4cc6055856e2. Accessed August 29, 2024.
proprietary modeling tool for multimodal access indicators, such as the tool developed by Conveyal. Proprietary tools also require financial resources as well as specialized expertise on staff.
Opportunity indicators as typically calculated also use geographic areas (e.g., census tracts or block groups) which, as noted in Chapter 3, are an imperfect proxy for populations of concern. A more accurate calculation of the benefits or burdens experienced by different population groups in a region would require allocating a percent of each census tract’s benefits or burdens to the population groups in the tract based on their share of that tract’s population. Again, this process, while an elaborate undertaking, can be automated with the use of scripts.
To increase data and metric access and use, community engagement is likely to be required to make analysis based on opportunity indicators meaningful. Engaging with underserved populations about destinations they value is a way to ground-truth the destinations used for proximity indicators and can provide input on culturally relevant destinations that might not be reflected in data sets with larger geographies. Engagement processes can also solicit the barriers individuals encounter, such as high-volume streets, inadequate pedestrian crossings, or infrequent transit service that similarly ground-truth the spatial analysis that produces opportunity metrics.
Finally, because travel demand models are already widely used for decision making in long-range plans, growth scenarios, needs assessments, and project prioritization, it makes sense that automating the calculation of opportunity indicators by demographic group and by mode in the most used travel demand models would increase the use of transportation equity measures by decision makers at the metropolitan level. These regional travel demand models are in public ownership and are a resource for state DOTs as well as MPOs and cover a majority of the nation’s population. Automating these calculations could be done by model application developers or by commissioning the writing of open-source scripts. Recognizing that proximity to destination metrics are a product of both infrastructure connections and the distribution of destinations, it is also important that forecasting models consider how land use and the location of destinations will change over time. This can be accomplished by incorporating land use forecasts into travel demand modeling scenarios or using integrated land use and travel demand models. Integrated land use and travel demand models are particularly well suited to this application because they can capture the dynamic interaction of transportation infrastructure investments on development patterns—for example, how a highway capacity project may reduce travel time but also enable more dispersed growth that decreases destination access over time.
TRANSPORTATION INSECURITY AS AN EQUITY METRIC
The development of tools to measure a person’s transportation insecurity has emerged from research on living with poverty. While definitions may vary slightly, in general, insecurity implies a lack of reliable and safe transportation options.14 To understand what it means to experience a lack of options, it is critical to interface through surveys or other qualitative methods with those most disadvantaged. While the proximity-to-destinations indicators described above can be used to analyze available opportunities, transportation insecurity addresses whether a person’s travel needs have been met. In this sense, transportation insecurity is complementary to proximity measures and can be used to identify less obvious transportation solutions. To assess transportation insecurity, University of Michigan researchers have developed the Transportation Security Index (TSI) measurement tools. Their approach to transportation insecurity takes its inspiration from decades of work at the national and international levels on food security, which provides models and lessons for operationalizing transportation insecurity on a large scale (see Box 4-1).15
There is no single source of data in the United States that can be used to systematically identify people for whom the transportation system is failing to meet their needs so severely that they are transportation insecure. Analyzing transportation security requires, as a necessary component, having data on missed trips or, to put it another way, on desired activities that do not occur because of transportation.16 Missed trips in transportation security are analogous to skipped meals in food security. Transportation insecurity also covers the transportation experience, seeking to identify people who are being forced to engage in transportation in such a way that it is injurious to their health or well-being. These types of questions are analogous in food security to questions about access to nutritious food.
___________________
14 USDOT defines transportation insecurity as occurring “when people are unable to get to where they need to go to meet the needs of their daily life regularly, reliably, and safely.” Researchers at the University of Michigan define transportation insecurity to be “the experience of being unable to move from place to place in a safe or timely manner.” (“Transportation Security Index.” Poverty Solutions, University of Michigan. https://poverty.umich.edu/research-funding-opportunities/data-tools/the-transportation-security-index. Accessed June 27, 2024.)
15 The terms transportation security or transportation insecurity used to define a lack of mobility is relatively new in transportation policy contexts, which may lead to some confusion with older uses of the term security to mean crime or terrorism affecting the transportation system.
16 USDOT’s ETC Explorer tool measures transportation insecurity through existing sources of socioeconomic and transportation data on transportation access (commute time, personal vehicle availability, and transit availability), transportation cost burden (percentage of household income spent on surface transportation), and transportation safety (fatality rate). The ETC Explorer’s measurement of transportation insecurity is not based on survey data of individuals’ transportation experiences and does not include missed trips.
BOX 4-1
Food Security Surveys as a Model for Measuring Transportation Security
The U.S. Department of Agriculture (USDA) defines “food security” for a household to mean “access by all members at all times to enough food for an active, healthy lifestyle.” This food must be “nutritionally adequate and safe,” and “access” means “the ability to acquire food in socially acceptable ways.” Food security excludes what USDA calls “coping strategies,” such as scavenging and stealing. USDA has developed a standard 10-question survey for adults, with additional questions for households with children. The survey does not ask respondents what they ate; instead, it asks them about skipped meals, cuts to portion sizes, and a perceived lack of nutritious food. Survey questions include whether respondents experienced worry or anxiety about losing access to food. USDA uses the survey tool to define households as food secure versus insecure (a binary category) and assign gradations of food insecurity (four categories, from least to most severe).a
USDA surveys Americans about their food security using two survey tools. The full survey is conducted annually, and a nationally representative sample of respondents are queried on their household’s food security experiences over the previous 12 months. The 2022 survey is based on data from almost 32,000 households, allowing comparisons by state.b Questions related to food security are also included in the Census Bureau’s Household Pulse Survey, which is an online survey on a wide range of topics that is deployed four times per year. The Household Pulse Survey asks respondents about the sufficiency of food over the past 7 days and includes a question on whether not having enough to eat was caused by not being able to get to a store because of transportation or mobility limitations.c
__________________
a U.S. Department of Agriculture (USDA). 2025. “Food Security in the U.S.: Measurement.” Updated October 25. https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-u-s/measurement.
b USDA conducts its annual survey of food security through the Food Security Supplement (FSS) of the monthly Current Population Survey (CPS), conducted by the U.S. Census Bureau for the Bureau of Labor Statistics. The FSS takes place as part of the December CPS and asks about food security over the past 12 months. See, for example, Rabbitt, M. P., L. J. Hales, M. P. Burke, and A. Coleman-Jensen. 2023. Household Food Security in the United States in 2022 (Report No. ERR-325). Economic Research Service, U.S. Department of Agriculture. https://doi.org/10.32747/2023.8134351.ers.
c The Household Pulse Survey is an experimental data product that was launched in April 2020. USDA. 2023. “Food Security in the United States: Documentation.” Updated October 25. https://www.ers.usda.gov/data-products/food-security-in-the-united-states/documentation/#hps; U.S Census Bureau. 2024. “Household Pulse Survey: Measuring Emergent Social and Economic Matters Facing U.S. Households.” April 2. https://www.census.gov/data/experimental-data-products/household-pulse-survey.html; U.S. Census Bureau. “Household Pulse Survey Phase 4.1.” https://www2.census.gov/programs-surveys/demo/technical-documentation/hhp/Phase_4-1_HPS_Questionnaire_English.pdf. Accessed June 27, 2024.
TSI and Uses for Equity Analysis
The TSI survey tools developed by researchers at the University of Michigan measure individual-level transportation insecurity as it is directly experienced, regardless of mode or geography.17 The tool is built around three “manifestations” or components of transportation security: material, relational, and emotional. Material manifestations of transportation insecurity include missed trips as well as trips that are excessively time consuming, such as having to arrive an hour early to an appointment or a job because of the bus schedule or the needs of the person providing the ride. Material manifestations also include unsafe trips. Relational manifestations of transportation insecurity cover strains on the transportation-insecure person’s relationships with others who they depend on for rides or who depend on their mobility. It also includes any resulting social isolation. Emotional manifestations of transportation insecurity are worry, anxiety, stress, and fear related to trip making or missed trips. This includes feeling “trapped” or “left out.”18
The TSI is currently available as a 16-question or 6-question survey module. Both versions have been validated as accurate at identifying people who are transportation (in)secure and for distinguishing levels or gradations of transportation security.19 Like the 10-question USDA food security
___________________
17 Alexandra K. Murphy, presentation to the committee, April 1, 2024. Although this chapter focuses on the TSI, researchers both domestic and international are working on developing similar survey tools for transportation insecurity. Please refer to Ng, A., D. Adjaye-Gbewonyo, and J. Dahlhamer. 2024. “Lack of Reliable Transportation for Daily Living, United States 2022.” National Center for Health Statistics, January 1. https://doi.org/10.15620/cdc:135611; Martens, K., and M. E. Singer 2024. “A Scale for Describing People’s Mobility Status.” Findings, April 19. https://doi.org/10.32866/001c.94195. See also Singer, M. E., and K. Martens. 2023. “Measuring Travel Problems: Testing a Novel Survey Tool in a Natural Experiment.” Transportation Research Part D: Transport and Environment 121(August):103834. https://doi.org/10.1016/j.trd.2023.103834. Pritchard, J. P., and K. Martens. 2023. “Towards Systematic Measurement of Travel Problems: A Pilot Study in the Greater Tel Aviv Area.” Travel Behaviour and Society 32(July):100591. https://doi.org/10.1016/j.tbs.2023.100591. Lättman, K., L. E. Olsson, and M. Friman. 2024. “Perceived Accessibility: Unveiling Inequalities in Transport Justice.” Sustainable Transport and Livability 1(1):2373050. https://doi.org/10.1080/29941849.2024.2373050. Lättman, K., L. E. Olsson, and M. Friman. 2016. “Development and Test of the Perceived Accessibility Scale (PAC) in Public Transport.” Journal of Transport Geography 54(June):257–263. https://doi.org/10.1016/j.jtrangeo.2016.06.015.
18 Alexandra K. Murphy, presentation to the committee, April 1, 2024. See also Gould-Werth, A., J. Griffin, and A. K. Murphy. 2018. “Developing a New Measure of Transportation Insecurity: An Exploratory Factor Analysis.” Survey Practice 11(2). https://doi.org/10.29115/SP-2018-0024.
19 Murphy, A. K., A. Gould-Werth, and J. Griffin. 2021. “Validating the Sixteen-Item Transportation Security Index in a Nationally Representative Sample: A Confirmatory Factor Analysis.” Survey Practice 14(1). https://doi.org/10.29115/SP-2021-0011. Murphy, A. K., A. Gould-Werth, and J. Griffin. 2024. “Using a Split-Ballot Design to Validate an Abbreviated Categorical Measurement Scale: An Illustration Using the Transportation Security Index.” Survey Practice 17(January). https://doi.org/10.29115/SP-2023-0030.
survey, the TSI can be used to distinguish the secure from the insecure and to measure levels of insecurity. Work is currently being done to establish an even shorter, three-question, valid survey tool. The questions included in the six-question TSI are listed in Box 4-2.
Testing the 16-item TSI survey also allowed the researchers to develop a first look at transportation insecurity in the United States. Survey results show that 25% of the population experiences some level of transportation insecurity, with 8% experiencing moderate or high levels. Of socioeconomic variables tested, the strongest correlate with transportation insecurity was poverty. Others more likely to experience transportation insecurity lacked a personal vehicle, were in the youngest age group (25–39), lived in cities, or lacked a high school diploma.20
The TSI, on its own, does not identify the consequences or the causes of transportation insecurity. This is by design, so that the TSI can be used as the independent or dependent variable in other research studies.21 For example, the TSI can be used as an independent variable in studies of health, employment, or educational outcomes.22 The TSI can also be used to evaluate the impact of transportation policies on transportation security. Examples to date include studies of bus passes, mobility wallets, and other policy interventions intended to reduce the cost of taking buses or taxis.23
Surveys of transportation insecurity could also be applied similarly to how USDA uses its surveys of food security (see Box 4-1): as a regular and appropriately frequent monitoring tool for transportation security in metropolitan, state, and national contexts. This application would fill an important gap in current knowledge regarding how well the transportation system is meeting the needs of all members of society. With a long enough time series of baseline data, such transportation security data could also be used to develop models that forecast the effects on transportation security—at the national, state, or regional level—of larger-scale transportation policies such as increases in the gas tax or vehicle registration fees or the implementation of congestion pricing.24
___________________
20 Murphy, A. K., K. McDonald-Lopez, N. Pilkauskas, and A. Gould-Werth. 2022. “Transportation Insecurity in the United States: A Descriptive Portrait.” Socius 8. https://doi.org/10.1177/23780231221121060; Rodier, C. 2024. “The Los Angeles Mobility Wallet: Results from a Randomized Control Trial Pilot Evaluation.” Conference on Advancing Transportation Equity, Baltimore, MD, July 17.
21 Alexandra K. Murphy, presentation to the committee, April 1, 2024.
22 The TSI is reportedly already being used in studies of medication adherence and health screening completion. Alexandra K. Murphy, presentation to the committee, April 1, 2024.
23 Alexandra K. Murphy, presentation to the committee, April 1, 2024.
24 Murphy outlined a taxonomy of ways the TSI is being or could be used in her presentation to the committee. Alexandra K. Murphy, presentation to the committee, April 1, 2024.
BOX 4-2
Transportation Security Index (TSI): Six-Item Survey Tool
The TSI, as developed by researchers at the University of Michigan, is a “validated measure of the experience of being unable to move from place to place in a safe or timely manner.” The six questions are as follows:
- To get to the places they need to go, people might walk, bike, take a bus, train or taxi, drive a car, or get a ride. In the past 30 days, how often did you have to reschedule an appointment because of a problem with transportation? [Often/Sometimes/Never]
- In the past 30 days, how often did you skip going somewhere because of a problem with transportation? [Often/Sometimes/Never]
- In the past 30 days, how often were you not able to leave the house when you wanted to because of a problem with transportation? [Often/Sometimes/Never]
- In the past 30 days, how often did you feel bad because you did not have the transportation you needed? [Often/Sometimes/Never]
- In the past 30 days, how often did you worry about inconveniencing your friends, family, or neighbors because you needed help with transportation? [Often/Sometimes/Never]
- In the past 30 days, how often did problems with transportation affect your relationships with others? [Often/Sometimes/Never]
Scoring each item is as follows: Never = 0, Sometimes = 1, Often = 2. Sum score 0–1: No insecurity/Secure; Sum score 2–5: Marginal/Low insecurity; Sum score 6–12: Moderate/High insecurity.
The TSI six-item survey tool also asks respondents how often they travel by mode, according to a list of 11 modes or ways of traveling.
SOURCE: “Administer the 6-Item Transportation Security Index.” Poverty Solutions, University of Michigan. https://sites.fordschool.umich.edu/poverty2021/files/2024/01/6-Item-Validated-Transportation-Security-Index.pdf. Accessed June 28, 2024.
The Minnesota Department of Transportation (MnDOT) has added the TSI’s six questions to its biennial “Omnibus Survey” of public opinion. The sampling strategy for this statewide survey only allows comparisons of two regions—the Twin Cities Metropolitan Area and Greater Minnesota—and, therefore, the survey results will not be useful for project planning and evaluation. However, the results will help MnDOT understand and track disparities in the state. For 2024, although the majority of respondents do not experience transportation insecurity, those with moderate or high insecurity are more likely to be younger, non-white, lower income, and less educated.
Preliminary cross tabulations based on race and ethnicity reveal moderate or high insecurity in historically disadvantaged communities. For non-Hispanic Blacks, non-Hispanic Asians, and non-Hispanic American Indian/Alaska Native, Native Hawaiian/Pacific Islanders, 27–36% experience moderate or high transportation insecurity compared to just 5% for non-Hispanic whites. MnDOT plans to further explore these data and to repeat these questions in future Omnibus surveys to establish trends. In addition, MnDOT is exploring using these questions in corridor planning and other public engagement efforts. If preproject information on transportation insecurity were available, postproject evaluations could be conducted.25
A measure such as the TSI is essential for expanding the usual focus on physical performance of infrastructure to understand whether transportation is meeting people’s needs. Such a measure can be applied to decisions about any mode or across modes. Several applications would be possible if agencies had TSI data with sufficient sample size and that were collected at fine enough spatial scales. Fine-scale geographic data, such as the census block group or tract, could be used to identify “hotspots” of insecurity.26 This information could be used by agencies to more effectively allocate transportation resources to improve accessibility, such as by creating new transit routes or stops, increase travel time reliability, for example by implementing high-occupancy vehicle toll lanes, and address traffic safety concerns by improving pedestrian and bicycle infrastructure, in addition to the location of social services. With regular surveys, transportation investments could be evaluated to determine if they moved people toward greater transportation security. Such surveys would be no more difficult to implement than surveys that collect national household travel data.
Notably, the transportation barriers or needs that people identify, using a measure such as the TSI, may be addressed by interventions other than investments in transportation infrastructure, for instance, by relocating critical services. Information about people’s transportation needs and barriers can encourage consideration of a broad range of options, including collaborations with agents outside of transportation in the search for solutions.
ENVIRONMENTAL JUSTICE AND PUBLIC HEALTH BURDENS
Public health and environmental justice burdens, which often impact individual as well as community health, are a broad category of equity concerns in transportation planning and decision-making processes. Three categories
___________________
25 Correspondence with Philip Schaffner, Director of the Office of Transportation System Management, MnDOT, August 28, 2024.
26 Murphy outlined a taxonomy of ways the TSI is being or could be used in her presentation to the committee. Alexandra K. Murphy, presentation to the committee, April 1, 2024.
of public health or environmental justice burdens are addressed below: transportation and housing cost burdens, transportation safety and security, and air quality. Cost burden metrics for housing and transportation bridge environmental justice concerns and access to housing, while also affecting access to destinations. In addition, incorporating housing and transportation cost burden metrics into decision making is an example of indicators facilitating analysis and projects across sectors. For transportation safety and personal security, reconceiving safety indicators according to the precepts of the safe system approaches and public health tools such as the Safe Systems Pyramid affects how these indicators—and potentially projects—are selected and deployed. In the air quality field, taking advantage of new, lower-cost technologies for data collection and analysis has the potential to dramatically improve understanding of disparate impacts and therefore enable the development of much more targeted solutions. Cost reductions in data and analysis technologies could have similar impacts for other environmental justice issues as well.
Emerging Metrics for Transportation and Housing Cost Burdens
Disparities in the cost of travel are an important indicator of inequitable transportation burdens. To more fully understand transportation’s role in the cost of living, most policymakers and advocates promote metrics that consider the cost of housing and transportation together. This is because areas with lower housing costs, which may appear more affordable for lower-income populations, may also have high transportation costs. The primary driver of a high cost of transportation is the ownership and operation of a motor vehicle or multiple motor vehicles per household. In areas with higher housing costs, households may be able to spend less on transportation, especially if the area has robust transit service and numerous places accessible by walking or rolling. Housing is typically considered affordable if it costs no more than 30% of household income and transportation if it costs no more than 15% of household income, for a combined housing plus transportation affordability threshold of below 45% of household income.27
___________________
27 To more fully describe housing affordability for lower-income households in a region, the standard method is to look at housing affordability at the 30% threshold for households with various percentages of the area’s median household income, such as 50% or 80% of median household income. Joice, P. 2014. “Measuring Household Affordability.” Cityscape: A Journal of Policy Development and Research 16(1):299–307. https://www.huduser.gov/portal/periodicals/cityscpe/vol16num1/ch17.pdf. For transportation affordability thresholds, see Hamidi, S., R. Ewing, and J. Renne. 2016. “How Affordable Is HUD Affordable Housing?” Housing Policy Debate 26(3):437–455. https://doi.org/10.1080/10511482.2015.1123753.
The Center for Neighborhood Technology’s (CNT’s) Housing + Transportation Affordability Index (H+T Index), which models housing and transportation costs nationwide, uses the 45% threshold.28 Building on CNT’s work, the U.S. Department of Housing and Urban Development (HUD) developed the Location Affordability Index (LAI) that models housing and transportation costs nationwide for eight different household types at the census-tract level. Five of the eight household types represent households below the area’s median household income.29 These two housing and transportation cost models were originally designed to inform decision making related to housing.
USDOT has begun to adopt metrics for transportation cost burden and housing cost burden but considers them separately. USDOT’s strategic plan currently includes among its equity key performance indicators to “reduce national transportation cost burden by 5%, including transportation travel cost as a percent of income by FY 2030.”30 USDOT’s “Equity Action Plan,” which also includes this performance indicator, finds that “inadequate coordination of land use, housing, and transportation policy and investment” can increase the transportation cost burden.31 Separate housing and transportation cost burden indicators are included in the USDOT ETC Explorer and are calculated nationwide at the census-tract level. The ETC Explorer’s transportation cost burden includes costs associated with vehicle ownership and operation, public transportation, and commuting; disadvantaged (burdened) status is based on percentile ranking (see the discussion of ETC Explorer in Chapter 3). The ETC Explorer estimates transportation costs using vehicle costs, transit costs, commute time, and a standardized value of time and drawing data from the Consumer Expenditure Survey, National Transit Database, and American Community Survey. The ETC Explorer’s housing cost burden indicator is the percent of households that spend over 30% of their income on housing.32
___________________
28 Center for Neighborhood Technology (CNT). “H+T Index.” https://htaindex.cnt.org. Accessed July 29, 2024.
29 U.S. Department of Housing and Urban Development (HUD). “Location Affordability Index.” https://www.hudexchange.info/programs/location-affordability-index. Accessed July 29, 2024; HUD. 2019. “Location Affordability Index: Version 3 Data and Methodology.” https://files.hudexchange.info/resources/documents/Location-Affordability-Index-Version-3-Data-and-Methodology.pdf.
30 U.S. Department of Transportation (USDOT). 2022. “Strategic Plan: FY 2022-2026.” https://www.transportation.gov/sites/dot.gov/files/2022-04/US_DOT_FY2022-26_Strategic_Plan.pdf.
31 USDOT. 2023. “Equity Action Plan: 2023 Update.” September. https://www.transportation.gov/sites/dot.gov/files/2023-12/2023%20update%20to%20the%20DOT%20Equity%20Action%20Plan.pdf.
32 USDOT. 2023. “ETC Explorer.” Updated December 4. https://www.transportation.gov/priorities/equity/justice40/etc-explorer.
Transportation agencies in states and metropolitan areas have also shown interest in transportation cost burdens, housing cost burdens, or both in formal transportation and housing metrics. MnDOT recently adopted transportation and housing cost as a performance measure. The measure uses cost-of-living data from the Minnesota Department of Employment and Economic Development and median household income from the American Community Survey to calculate an income threshold needed for housing and transportation needs. This value is estimated at a statewide level for a typical household of two adults and one child. In the future, MnDOT plans to report this measure by county. Although a smaller unit of geography is desirable, the data sources available annually are only by county. MnDOT is in the process of developing targets based on this measure. The performance measure dashboard does not currently track transportation and housing cost burdens for households below the median income.33
Several statewide transportation plans describe goals to reduce transportation cost burdens, housing cost burdens, or both, including the Oregon Transportation Plan (2023), the California Transportation Plan 2050, and the 2045 Hawaii Statewide Transportation Plan. These measures are in the early stages of being operationalized, with no thresholds or targets set and limited information about how they guide investment and planning decisions.34
Uses of Cost Metrics for Equity Analysis
The housing and transportation cost burden metric holistically assesses the contribution of land use and transportation to outcomes related to income and wealth. The metric reflects the relationship between access and housing affordability and can identify opportunities where transportation and housing policy can collaborate on equity-focused solutions. In one example of its use, a nonprofit housing group partnered with university researchers to create and evaluate an electric car sharing service located at affordable housing developments. The service enabled residents to access more
___________________
33 MnDOT. “Performance Measure Dashboard: Cost of Transportation and Housing.” https://www.dot.state.mn.us/measures/cost-transportation-housing.html. Accessed July 30, 2024.
34 The Caltrans “transportation and housing cost burden” performance measure is to be calculated by income quintile and race. Caltrans. 2021. “California Transportation Plan 2050.” February. https://dot.ca.gov/-/media/dot-media/programs/transportation-planning/documents/ctp-2050-v3-a11y.pdf; Hawaii DOT. “Hawaii Statewide Transportation Plan.” https://hidot.hawaii.gov/administration/hawaii-statewide-transportation-plan. Accessed July 30, 2024; Oregon DOT. 2023. “Oregon Transportation Plan.” July 13. https://www.oregon.gov/odot/Planning/Pages/Oregon-Transportation-Plan-Update.aspx.
opportunities while providing a low-cost alternative to car ownership. The service has since spun off as a nonprofit organization and has expanded to additional affordable housing developments in the region.35
The Metropolitan Transportation Commission in the San Francisco Bay Area uses rent-burdened households as one of the indicators to define equity priority communities. This variable is defined as the share of households whose expenses for rent exceed 50% of their income. When this value exceeds a particular threshold (14%), the census tract is considered to have a high share of rent-burdened households.36
Another example of collaborating across domains and combining metrics to assess transportation equity is the pilot implementation of a mobility wallet in Los Angeles, administered by Metro. The wallet provides a monthly stipend to low-income individuals registered in the program for spending on clean transportation options, like transit, ride-hailing, and bikeshare services. Researchers are evaluating the program for its effects on access, stress, and contributions to transportation security using a version of the TSI described earlier.37
These examples emphasize how the transportation and housing cost burden considered together can compound to create more equitable outcomes and minimize cost burdens. Multidimensional metrics that capture the multidimensional nature of inequity provide support for innovative solutions within and beyond transportation.
Data, Analysis, and Capacity Needs
Both the H+T Index and the LAI have peer-reviewed methodologies for calculating housing and transportation costs. Both indices use aggregate data to model costs over census geographies from a limited set of variables.38 Although these methods provide comprehensive coverage using readily available data via the American Community Survey and, in the case of
___________________
35 Harold, B., C. Rodier, and Y. Zhang. 2022. “Retrospective User Survey for a Rural Electric Vehicle Carsharing Pilot in California’s Central Valley.” University of California Institute of Transportation Studies, UC Office of the President. https://doi.org/10.7922/G2CJ8BT5.
36 Metropolitan Transportation Commission (MTC). 2024. “Equity Priority Communities.” Updated May 20. https://mtc.ca.gov/planning/transportation/access-equity-mobility/equity-priority-communities.
37 Rodier, C. 2024. “The Los Angeles Mobility Wallet: Results from a Randomized Control Trial Pilot Evaluation.” Conference on Advancing Transportation Equity, Baltimore, MD, July 17.
38 Haas, P. M., G. L. Newmark, and T. R. Morrison. 2016. “Untangling Housing Cost and Transportation Interactions: The Location Affordability Index Model—Version 2 (LAIM2).” Housing Policy Debate 26(4–5):568–582. https://doi.org/10.1080/10511482.2016.1158199; Hass, P. M., C. Makarewicz, A. Benedict, and S. Bernstein. 2008. “Estimating Transportation Costs by Characteristics of Neighborhood and Household.” Transportation Research Record 2077(1):62–70. https://doi.org/10.3141/2077-09.
the H+T Index, additional data based on vehicle odometer readings from a limited number of states, they have been criticized for relying on aggregated data and unreliable estimates of transit use.39 Aggregated data masks variation in household-level characteristics, such as vehicle ownership and income, with the potential to miss burdened households in nonburdened tracts. The indices also omit certain variables, like race, that some have found to be independently associated with transportation costs.40
Another missing element from these transportation cost metrics are accurate estimates of the time cost of travel, in addition to the total cost of travel. An accurate time cost of travel is important because it reflects a trade-off between housing price and access to destinations. Places with higher housing costs are likely to also have higher access to destinations and thus can reach more destinations in less time. While USDOT’s definition of transportation cost burden in the ETC Explorer includes a commuting cost, the cost is given as a constant by trip purpose or mode and does not vary for household characteristics. This can be problematic because using a single monetary valuation for travel time, as done in traditional cost-benefit analysis, has been criticized for producing higher valuations for investments that favor higher-income transportation users.41 Reduced and reliable travel times may be more valuable to lower-income individuals given that they often experience severe constraints on their schedules, such as the need to carefully plan travel to multiple places, like work and daycare, in the context of fixed transit schedules and employer-imposed penalties for arriving late to work. In any case, a more comprehensive accounting of the time cost of travel will shift the transportation calculation of the housing and
___________________
39 The most recent version of HUD’s Location Affordability Index is from 2019. Although many of the critiques cited here reviewed an earlier version of the LAI, the critiques also apply to the 2019 version. Hamidi, S., R. Ewing, and J. Renne. 2016. “How Affordable Is HUD Affordable Housing?” Housing Policy Debate 26(3):437–455. https://doi.org/10.1080/10511482.2015.1123753; Ganning, J. P. 2017. “It’s Good but Is It Right? An Under-the-Hood View of the Location Affordability Index.” Housing Policy Debate 27(6):807–824. https://doi.org/10.1080/10511482.2017.1312478; Smart, M. J., and N. J. Klein. 2018. “Complicating the Story of Location Affordability.” Housing Policy Debate 28(3):393–410. https://doi.org/10.1080/10511482.2017.1371784; Makarewicz, C., P. Dantzler, and A. Adkins. 2020. “Another Look at Location Affordability: Understanding the Detailed Effects of Income and Urban Form on Housing and Transportation Expenditures.” Housing Policy Debate 30(6):1033–1055. https://doi.org/10.1080/10511482.2020.1792528.
40 Renne, J. L., T. Tolford, S. Hamidi, and R. Ewing. 2016. “The Cost and Affordability Paradox of Transit-Oriented Development: A Comparison of Housing and Transportation Costs Across Transit-Oriented Development, Hybrid and Transit-Adjacent Development Station Typologies.” Housing Policy Debate 26(4–5):819–834. https://doi.org/10.1080/10511482.2016.1193038.
41 Martens, K., and F. Di Ciommo. 2017. “Travel Time Savings, Accessibility Gains and Equity Effects in Cost–Benefit Analysis.” Transport Reviews 37(2):152–169. https://doi.org/10.1080/01441647.2016.1276642.
transportation cost burden and likely demonstrate a higher transportation cost burden in more sprawling areas.
As with all indices that rely on aggregate data, the calculation of housing and transportation cost burden would be more robust if individual-level and local data were available. Specific information on trip distance by mode, calculated according to demographic characteristics like income, would give a more accurate picture of costs incurred by individual travelers. And as others have shown, disaggregating these data by race and ethnicity and other individual characteristics is important for proposing equitable remedies. National or regional household travel surveys can be used to gather some of this information, supplemented by passively collected traveler data sets such as mobile phone location data.
Longitudinal data on housing and transportation costs are important to track trends over time. Much of the recent research on transportation and urban form (how the physical layout of a city affects the movement of people and goods) and their impacts on household transportation and housing expenditures comes from the Panel Survey of Income Dynamics (PSID), a rich longitudinal survey that provides detailed estimates of transportation costs at a fine spatial resolution.42 Although this survey comes close to representing the U.S. population, it is national in scope and oversamples suburban areas, leaving some bias or nonrepresentation for small-area estimates of other locations.43 Adding questions on housing expenditures to transportation surveys would help supplement this data gap. Despite data gaps, the longitudinal nature of the PSID is an important asset and could be a model for collecting data specific to the links between housing, transportation, and employment that lead to changes in wealth outcomes over time.
___________________
42 For recent research using the PSID, see Smart, M. J., and N. J. Klein. 2017. “Complicating the Story of Location Affordability.” Housing Policy Debate 28(3):393–410. https://doi.org/10.1080/10511482.2017.1371784; Makarewicz, C., P. Dantzler, and A. Adkins. 2020. “Another Look at Location Affordability: Understanding the Detailed Effects of Income and Urban Form on Housing and Transportation Expenditures.” Housing Policy Debate 30(6):1033–1055. https://doi.org/10.1080/10511482.2020.1792528; Schouten, A. 2022. “Residential Location and Household Spending: Exploring the Relationship Between Neighborhood Characteristics and Transportation and Housing Costs.” Urban Affairs Review 58(6):1554–1584. https://doi.org/10.1177/10780874211028814; Molloy, Q., N. Garrick, and C. Atkinson-Palombo. 2024. “Black Households Are More Burdened by Vehicle Ownership Than White Households.” Transportation Research Record 2678(10):163–173. https://doi.org/10.1177/03611981241231968.
43 Makarewicz, C., P. Dantzler, and A. Adkins. 2020. “Another Look at Location Affordability: Understanding the Detailed Effects of Income and Urban Form on Housing and Transportation Expenditures.” Housing Policy Debate 30(6):1033–1055. https://doi.org/10.1080/10511482.2020.1792528.
Transportation Safety and Personal Security
Deaths and serious injuries on the roads form one of the most significant environmental justice burdens from the transportation system. As described in Chapter 2, traffic crashes and fatalities have worsened over the past decade, and much of the burden falls on people of color. In response to the continuing and recently worsening road safety crisis in the United States, many local governments have adopted a safe system approach to roadway safety such as Vision Zero. With the adoption of USDOT’s National Roadway Safety Strategy (NRSS) in January 2022, a safe system approach is official federal policy.44 These safe system approaches are replacing a dominant safety paradigm built around education, engineering, and traffic enforcement (the “3 Es”) that focused much of its effort on changing human behavior and put primacy on the safety of those inside vehicles.45 All safe system approaches include data-driven analysis to identify and prioritize the most dangerous intersections and corridors for safety interventions. However, one critique of typical safe system approaches is that although they broaden the field of action—adding “Es” such as emergency response—they often do not adequately address that not every “E” component is equally effective at reducing harm nor implicated in the root causes of fatalities and serious injuries.46
Public Health and Safe System Approaches
The Safe Systems Pyramid for traffic safety (see Figure 4-1), as introduced by David J. Ederer and colleagues, brings a stronger public health orientation into safe system approaches for road safety. The Pyramid is modeled on a conceptual framework used for efficacy in public health interventions, adopted by the U.S. Centers for Disease Control and Prevention (CDC) and other organizations focused on achieving positive public health
___________________
44 Vision Zero Network. “Vision Zero Communities.” https://visionzeronetwork.org/resources/vision-zero-communities. Accessed July 3, 2024; USDOT. 2024. “2024 Progress Report on the National Roadway Safety Strategy.” Updated May 1. https://www.transportation.gov/nrss/2024-progress-report-national-roadway-safety-strategy.
45 Dumbaugh, E., and J. L. Gattis. 2005. “Safe Streets, Livable Streets.” Journal of the American Planning Association 71(3):283–300. https://doi.org/10.1080/01944360508976699; Dumbaugh, E., D. Saha, and L. Merlin. 2024. “Toward Safe Systems: Traffic Safety, Cognition, and the Built Environment.” Journal of Planning Education and Research 44(1):75–87. https://doi.org/10.1177/0739456X20931915.
46 Ederer, D. J. 2023. “The Safe Systems Pyramid: A New Framework for Traffic Safety.” Transportation Research Interdisciplinary Perspectives 21. https://doi.org/10.1016/j.trip.2023.100905. For example, the NRSS adopts safe system principles but still organizes itself around the remnants of the “3 Es”: safer people, safer roads, safer vehicles, safer speeds, and post-crash care. Except for “safer speeds,” the other four are variations on “Es.”
SOURCE: Ederer, D. J., R. Panik, N. Botchwey, and K. Watkins. 2023. “The Safe Systems Pyramid: A New Framework for Traffic Safety.” Transportation Research Interdisciplinary Perspectives 21. https://doi.org/10.1016/j.trip.2023.100905.CC-BY-NC-ND.
outcomes.47 The Pyramid recognizes five types of interventions organized along two continuums that trade off individual effort and population health impact. The Pyramid prioritizes those nearer the base of the pyramid (high population health impact/low individual effort) because they are likely to be the most effective interventions. This prioritization of low-individual-effort interventions is a general rule: certain specific interventions higher up in the pyramid may have greater effectiveness than more foundational interventions. The top of the pyramid is Education, the category with the highest requirements for individual effort. Active Measures still require the traveler to act, such as stopping at stop signs, crossing only at a walk light, or putting on a seat belt. Latent Measures, because they do not require individual action, are likely to be highly effective, but they still apply at the individual level. Examples include air bags, leading pedestrian intervals (a traffic signal that gives pedestrians a head start when crossing the street), and automated traffic law enforcement.
___________________
47 Frieden, T. R. 2010. “A Framework for Public Health Action: The Health Impact Pyramid.” American Journal of Public Health 100(4):590–595. https://doi.org/10.2105/AJPH.2009.185652.
The Built Environment category includes engineering approaches designed to reduce the transfer of energy during a crash or the population’s exposure to dangerous situations. This category also includes land use, population density, and access-to-destinations interventions that reduce exposure to safety risks. The Socioeconomic Factors category includes the socioeconomic determinants of health more broadly. These are the social and economic factors that influence people’s need to travel and when, where, and how they travel. This includes factors like working the night shift, when driving is more dangerous and mass transit less available. It also includes geographic disparities in traffic safety risk. For example, in communities where modest apartment complexes are concentrated along wide, high-traffic commercial arterials, people with lower incomes may be at a higher risk. Table 4-1 has additional examples of transportation-related indicators for the factor types in the Safe Systems Pyramid.
The key difference between the Safe Systems Pyramid and more typical safe system approaches is that the Pyramid organizes and pursues interventions not by topic area, but by the type of action or the size of the public health impact. The Safe Systems Pyramid also emphasizes certain built environment or socioeconomic factors that may be difficult for a transportation agency to address on its own. Regardless of the type of safe system
TABLE 4-1 Illustrative Measures for the Levels of the Safe Systems Pyramid
| Factor | Illustrative Indicators for Metrics |
|---|---|
| Socioeconomic factors | Community income mix, percent of homes affordable for low-income individuals |
| Built environment | Walkability index, bikeability index/level of traffic stress, protected intersections installed (peds and bikes), residential density, land use mix, percent complete streets, percent roadways with <25 mph limit, lighting |
| Latent safety measures | Vehicles with pedestrian crash worthiness, percent cars sold with speed governors, interlock devices, distraction prevention; roadway lighting, leading pedestrian intervals, automatic pedestrian signals (e.g., High-Intensity Activated Crosswalk [HAWK] signals) |
| Active safety measures | Percent cars with pedestrian warning alerts, compliance with posted signs and signals |
| Education | Number and effectiveness of driver education campaigns |
SOURCE: Committee analysis to supplement Ederer, D. J., R. Panik, N. Botchwey, and K. Watkins. 2023. “The Safe Systems Pyramid: A New Framework for Traffic Safety.” Transportation Research Interdisciplinary Perspectives 21. https://doi.org/10.1016/j.trip.2023.100905.CC-BY-NC-ND.
approach, to apply such approaches in the United States requires improved data collection and analysis methods. Moreover, these improved data and analysis methods would need to be able to be incorporated into equity analysis. Community-informed data collection and analysis methods would be part of the solution.
The safe system approach and the Safe Systems Pyramid as depicted in Figure 4-1 focus on the interventions that reduce traffic crashes and mitigate their effects. However, traffic safety is just one element of safety and personal security in the transportation system. Personal identity characteristics, sociopsychological factors, and environmental characteristics create real and perceived vulnerabilities that affect whether and how people get around.48 For example, in response to harassment on public transportation, women and gender minorities modify their behavior to take preventive measures, including changing their mode preference.49 For girls of color ages 12–15, their own safety precautions together with societal expectations to stay at home discourages using active transportation.50 The incorporation of police enforcement within the safe system approach, even with emphasis on ensuring equity in enforcement, offers opportunities for racial bias in stops, searches, and citations to permeate the transportation system.51 These inequities cause changes in travel behavior and have consequences beyond transportation and travel themselves.52 An opportunity
___________________
48 Loukaitou-Sideris, A. 2006. “Is It Safe to Walk? Neighborhood Safety and Security Considerations and Their Effects on Walking.” Journal of Planning Literature 20(3):219–232. https://doi.org/10.1177/0885412205282770.
49 Loukaitou-Sideris, A. 2014. “Fear and Safety in Transit Environments from the Women’s Perspective.” Security Journal 27(2):242–256. https://doi.org/10.1057/sj.2014.9; Loukaitou-Sideris, A. 2016. “A Gendered View of Mobility and Transport: Next Steps and Future Directions.” Town Planning Review 87(5):547–565. https://doi.org/10.3828/tpr.2016.38; Loukaitou-Sideris, A., M. Brozen, M. Pinski, and H. Ding. 2024. “Documenting #MeToo in Public Transportation: Sexual Harassment Experiences of University Students in Los Angeles.” Journal of Planning Education and Research 44(1):210–224. https://doi.org/10.1177/0739456X20960778.
50 Roberts, J. D., S. Mandic, C. S. Fryer, M. L. Brachman, and R. Ray. 2019. “Between Privilege and Oppression: An Intersectional Analysis of Active Transportation Experiences Among Washington D.C. Area Youth.” International Journal of Environmental Research and Public Health 16(8):1313. https://doi.org/10.3390/ijerph16081313.
51 Pierson, E., C. Simoiu, J. Overgoor, S. Corbett-Davies, D. Jenson, A. Shoemaker, V Ramachandran, P. Barghouty, C. Phillips, R. Shroff, and S. Goel. 2020. “A Large-Scale Analysis of Racial Disparities in Police Stops Across the United States.” Nature Human Behaviour 4(7):736–745. https://doi.org/10.1038/s41562-020-0858-1; Barajas, J. M. 2021. “Biking Where Black: Connecting Transportation Planning and Infrastructure to Disproportionate Policing.” Transportation Research Part D: Transport and Environment 99:103027. https://doi.org/10.1016/j.trd.2021.103027.
52 Baumgartner, F. R., D. A. Epp, and K. Shoub. 2018. Suspect Citizens: What 20 Million Traffic Stops Tell Us About Policing and Race. New York: Cambridge University Press. https://public.ebookcentral.proquest.com/choice/publicfullrecord.aspx?p=5473139.
exists to extend the framework proposed by the Safe Systems Pyramid to these and other factors that take a holistic approach to safety and security in the transportation system.
Methods to Improve Equity in Safety Data
Methods to improve equity in safety data can focus on efforts to enhance crash reporting and analysis and efforts to incorporate additional safety and security considerations into data analysis. First, safe system approaches require accurate data on crash characteristics. However, most crash data in the United States are collected via police reports, and studies have shown that a substantial share of reports have inaccuracies that can lead to erroneous conclusions.53 Moreover, police reports cover existing conditions and road user behavior but are not designed to determine the root causes of crashes, making them inadequate for safe system approaches.54
Crashes of all kinds are underreported, and the prevalence of underreporting increases as the severity of injury decreases.55 This underreporting makes it difficult to detect a pattern of smaller crashes before the big crash that kills people. Missing data often contain bias as well; for example, crashes injuring Black individuals are less likely to result in a police report, leaving these crashes out of official statistics.56 Underreporting of crashes
___________________
53 Yamamoto, T., J. Hashiji, and V. N. Shankar. 2008. “Underreporting in Traffic Accident Data, Bias in Parameters and the Structure of Injury Severity Models.” Accident Analysis & Prevention 40(4):1320–1329. https://doi.org/10.1016/j.aap.2007.10.016; Burdett, B., Z. Li, A. R. Bill, and D. A. Noyce. 2015. “Accuracy of Injury Severity Ratings on Police Crash Reports.” Transportation Research Record 2516(1):58–67. https://doi.org/10.3141/2516-09.
54 Ederer, D. J., R. Panik, N. Botchwey, and K. Watkins. 2023. “The Safe Systems Pyramid: A New Framework for Traffic Safety.” Transportation Research Interdisciplinary Perspectives 21. https://doi.org/10.1016/j.trip.2023.100905.
55 Elvik, R., and A. Mysen. 1999. “Incomplete Accident Reporting: Meta-Analysis of Studies Made in 13 Countries.” Transportation Research Record 1665(1):133–140. https://doi.org/10.3141/1665-18; Winters, M., and M. Branion-Calles. 2017. “Cycling Safety: Quantifying the Under Reporting of Cycling Incidents in Vancouver, British Columbia.” Journal of Transport & Health, Road Danger Reduction 7:48–53. https://doi.org/10.1016/j.jth.2017.02.010; Medury, A., O. Grembek, and A. Loukaitou-Sideris, K. Shafizadeh. 2019. “Investigating the Underreporting of Pedestrian and Bicycle Crashes in and Around University Campuses—a Crowdsourcing Approach.” Accident Analysis & Prevention 130:99–107. https://doi.org/10.1016/j.aap.2017.08.014; Imprialou, M., and M. Quddus. 2019. “Crash Data Quality for Road Safety Research: Current State and Future Directions.” Accident Analysis & Prevention, Road Safety Data Considerations 130:84–90. https://doi.org/10.1016/j.aap.2017.02.022.
56 Sciortino, S., M. Vassar, M. Radetsky, and M. Knudson. 2005. “San Francisco Pedestrian Injury Surveillance: Mapping, Under-Reporting, and Injury Severity in Police and Hospital Records.” Accident Analysis & Prevention 37(6):1102–1113. https://doi.org/10.1016/j.aap.2005.06.010; Edwards, M., and M. Gutierrez. 2023. “The Incidence Burden of Unreported Pedestrian Crashes in Illinois.” Traffic Injury Prevention 24(1):82–88. https://doi.org/10.1080/15389588.2022.2143236.
involving people of color or that occurred in Black neighborhoods can lead to fewer or mistargeted safety interventions, thereby perpetuating racial disparities in crash outcomes.
Linking hospital records to police report data is one way to get a more comprehensive and equitable perspective on crash injuries. Research finds that hospital records typically include crashes not found in police databases. In addition, the crash types and personal characteristics in hospital records differ from those found in police reports; hospital records tend to have a wider range of injury types and severity and more socioeconomic diversity among victims. The National Syndromic Surveillance Program (NSSP) of CDC is a source of near-real-time data from emergency departments across the United States.57 The NSSP has been used to demonstrate persistent racial disparities in pedestrian injuries and could be an important source of data for local conditions.58 Because the system is not limited to public roads, CDC’s Web-based Injury Statistics Query and Reporting System (WISQARS) database could also be used to examine disparities in fatal and nonfatal injuries, while encompassing a wider set of locations than are available in traditional police reports. However, WISQARS only provides data at the county level.59 Extending these databases to be more accurate for real-time reporting may help in developing effective safety interventions.
Crowdsourcing is another way to develop a more comprehensive data set of crashes and safety hazards that include the minor injury, single-person, and near-miss events that are most likely to never make it into police reports. Crowdsourced data are volunteered geographic information on the location of safety incidents, reported by the people involved or witnesses to an online platform designed for this purpose. Crowdsourced data fill gaps in spatial information on potentially dangerous locations that would otherwise go unacknowledged. Crowdsourcing platforms may also help to mitigate racial bias in traffic crash reporting, by allowing users to
___________________
57 U.S. Centers for Disease Control and Prevention (CDC). “National Syndromic Surveillance Program (NSSP).” https://www.cdc.gov/nssp/index.html.
58 Barry, V., M. Van Dyke, J. Nakayama, H. Zaganjor, M. Sheppard, Z. Stein, L. Radhakrishnan, E. Schweninger, K. Rose, G. Whitfield, and B. West. 2024. “Emergency Department Visits for Pedestrians Injured in Motor Vehicle Traffic Crashes—United States, January 2021–December 2023.” Morbidity and Mortality Weekly Report 73(17):387–392. https://doi.org/10.15585/mmwr.mm7317a1.
59 CDC. “WISQARS: Web-based Injury Statistics Query and Reporting System.” https://wisqars.cdc.gov.
report safety incidents without using the police as an intermediary.60 However, crowdsourcing platforms require wide and constant advertisement to be useful. Even with such publicity, the data collected may be biased toward those who can readily access web-based platforms or otherwise have a strong interest in reporting (e.g., avid cyclists). These kinds of tools also permit capture of data that contribute to the overall safety and security experience. For example, early analysis of a pedestrian-focused tool showed that safety and comfort concerns and missing benches were among the top reasons reported for why pedestrians did not feel safe in a neighborhood or along a street segment while walking.61 Quantitative data such as street lighting conditions and incidents of harassment, and qualitative data such as perceptions of safety and characteristics of near misses can be collected to develop a holistic picture of safety.
A third method is to combine volunteered geographic information with community engagement. Street Story is a community engagement tool developed by the Safe Transportation Research and Education Center at the University of California, Berkeley, that can be used to gather safety data. These data and the mapping tool are designed for “residents, organizations, or agencies to collect local information about traffic crashes, near-misses, general hazards and safe locations to travel.”62 Walk audits are a similar tool, targeted at pedestrian safety and promoting pedestrian-friendly infrastructure. Walk audits may be relatively low tech—they do not require an online mapping tool—and typically rely on volunteers to assess neighborhood safety.63 As community engagement tools, their output of these and similar tools is designed to be useful for local transportation planning or policy development. They are not designed to produce consistent data across communities and regions.
___________________
60 See, for example, “BikeMaps.” https://bikemaps.org. Accessed July 1, 2024; Fischer, J., S. Sersli, T. Nelson, H. Yu, K. Laberee, M. Zanotto, and M. Winters. 2022. “Spatial Variation in Bicycling Risk Based on Crowdsourced Safety Data.” Canadian Geographies/Géographies Canadiennes 66(3):556–568. https://doi.org/10.1111/cag.12756; Ferster, C. J., T. Nelson, M. Winters, and K. Laberee. 2017. “Geographic Age and Gender Representation in Volunteered Cycling Safety Data: A Case Study of BikeMaps.Org.” Applied Geography 88(November 1):144–150. https://doi.org/10.1016/j.apgeog.2017.09.007.
61 Laberee, K., T. Nelson, D. Boss, C. Ferster, K. Hosford, D. Fuller, M-S. Cloutier, and M. Winters. 2023. “WalkRollMap.Org: Crowdsourcing Barriers to Mobility.” Frontiers in Rehabilitation Sciences 4. https://doi.org/10.3389/fresc.2023.1023582.
62 Berkeley SafeTREC. 2021. “Street Story: Starter Guide for Communities and Agencies.” https://streetstory.berkeley.edu/assets/Street%20Story_Starter%20Guide%20for%20Communities%20and%20Agencies_20211213.pdf.
63 For examples of walk audit tools, refer to Safe Routes to School National Partnership. 2018. “Let’s Go for a Walk: A Toolkit for Planning and Conducting a Walk Audit.” January. https://www.saferoutespartnership.org/sites/default/files/walk_audit_toolkit_2018.pdf; “AARP Walk Audit Tool Kit.” Updated June 2022. https://www.aarp.org/livable-communities/getting-around/aarp-walk-audit-tool-kit.html.
Finally, a long-standing challenge in properly characterizing inequitable risk on the road system is a lack of accurate exposure data, especially for walking and cycling. Exposure refers to the amount of travel done or activities participated in, and exposure metrics are used as the denominator in the assessment of risk.64 For the active modes, exposure is often measured through manual or automatic counters and survey-based estimates, such as those from regional or national travel surveys, the American Community Survey, or other purpose-built surveys.65 Each of these methods has their drawbacks; counters are limited in space, while surveys are limited in time and may not capture all travel purposes, yielding different estimates of crash risk.66 Proprietary data sources aim to fill these gaps using mobile phone location data, user-contributed activity data, and modeled volumes, which cities are increasingly using for planning purposes. These data sources also have limitations in their demographic representativeness: crowdsourced fitness app data, for example, have documented biases toward young, male, white bicyclists.67 Statistical methods to correct such bias that draw on easily accessible data sources are emerging, but these methods must be combined with community engagement approaches to improve the representativeness of the contributed data themselves.68
Evaluating Equity in Safety Decision Making
Evaluating equity in transportation safety programs requires data and methods for comparative analysis that establish whether current safety investments are—or are not—going to populations experiencing disproportionate harm. Such comparative analysis can take advantage of existing performance measure processes at the state and local levels. The federal safety performance management (Safety PM) rule requires state DOTs and
___________________
64 Merlin, L. A., E. Guerra, and E. Dumbaugh. 2020. “Crash Risk, Crash Exposure, and the Built Environment: A Conceptual Review.” Accident Analysis & Prevention 134:105244. https://doi.org/10.1016/j.aap.2019.07.020.
65 Ferenchak, N. N., and W. E. Marshall. 2020. “Is Bicycling Getting Safer? Bicycle Fatality Rates (1985–2017) Using Four Exposure Metrics.” 100219. https://doi.org/10.1016/j.trip.2020.100219.
66 Ibid.
67 Lee, K., and I. N. Sener. 2021. “Strava Metro Data for Bicycle Monitoring: A Literature Review.” Transport Reviews 41(1):27–47. https://doi.org/10.1080/01441647.2020.1798558.
68 Roy, A., T. Nelson, A. Fotheringham, and M. Winters. 2019. “Correcting Bias in Crowdsourced Data to Map Bicycle Ridership of All Bicyclists.” Urban Science 3(2):62. https://doi.org/10.3390/urbansci3020062; Garber, M. D., K. E. Watkins, and M. R. Kramer. 2019. “Comparing Bicyclists Who Use Smartphone Apps to Record Rides with Those Who Do Not: Implications for Representativeness and Selection Bias.” Journal of Transport & Health 15:100661. https://doi.org/10.1016/j.jth.2019.100661.
MPOs to make publicly available data on five safety performance measures for their public roads:
- Number of Fatalities
- Rate of Fatalities per 100 Million Vehicle Miles Traveled (VMT)
- Number of Serious Injuries
- Rate of Serious Injuries per 100 Million VMT
- Number of Non-Motorized Fatalities and Non-Motorized Serious Injuries69
These measures are reviewed annually by the decision makers responsible for allocating funding resources.70 To evaluate equity, they can be calculated for different demographic groups or for subareas with a high percentage of disadvantaged populations. Safety equity performance measure(s) are not required under the Safety PM rule, but some DOTs and MPOs evaluate demographic differences periodically as part of their safe systems approach.
The distribution of the crash burden across demographic groups can be combined with the reviews of the distribution of transportation investments discussed in Chapter 3. For example, the Mid-American Regional Council (MARC) conducted a geographic comparative analysis for pedestrian fatalities along with a geographic distributive analysis of its funding. They found that pedestrian fatalities were highest in underserved communities, but better resourced communities were receiving a disproportionate share of the safety funding. They determined that the root cause of the disparity in funding was a grant application process that was easier for better-resourced communities to access. MARC responded by directing additional resources to underserved communities.71
Traffic-Related Air Pollution
Motor vehicle traffic is a large source of local and regional air pollution in the United States. Exposure to traffic-related air pollution (TRAP) is correlated with adverse health outcomes, including heart disease, lung cancer, and asthma, as discussed in Chapter 2. Motor vehicle traffic contributes to both local air pollution hotspots along roadways and to regional air pollution concerns such as ozone and smog. Local exposure hotspots formed from “primary” or directly emitted air pollutants such as particulate matter
___________________
69 The committee is not implying these are sufficient in all contexts.
70 FHWA (Federal Highway Administration). 2022. “Safety Performance Management (Safety PM).” Updated June 28. https://safety.fhwa.dot.gov/hsip/spm.
71 Equity Analysis in Regional Transportation Planning Processes, Volume 2: Research Overview, Transit Cooperative Research Program (TCRP) 214, 2020, 27.
(which include emissions from tire and brake wear) and nitrogen dioxide that form near high concentrations of vehicle traffic have been the focus of equity concerns and efforts to measure air quality impacts.
The Clean Air Act directs the U.S. Environmental Protection Agency (EPA) to set air quality standards that protect health and to regulate emissions from vehicles. To fulfill its mandate, EPA has developed what are by now well-established tools and methods for estimating the air quality effects of transportation projects and to a larger extent regional transportation plans. One way that EPA works to minimize exposure to air pollution is through the National Ambient Air Quality Standards (NAAQS), which define the maximum allowable ambient concentrations for six criteria air pollutants. The primary NAAQS are set at levels designed to protect public health. Regions of the United States that do not meet these standards are classified as Non-Attainment Areas.
Advances in mobile source emission and air quality analysis methods are more commonly deployed in Non-Attainment Areas, where more stringent air quality analysis is often required. These advances could be adopted more widely and applied to project-level or neighborhood-level TRAP impact analysis, including to identify environmental justice “hotspots” and evaluate project-level impacts and mitigation strategies. They could also be applied during planning stages to analyze transportation and land use scenarios and project alternatives. In addition, the adoption of easy-to-use, low-cost air quality sensors can develop more granular data on air quality, while empowering local governments and community groups. In the absence of advanced air quality modeling, there are also advances in simpler techniques to identify TRAP-impacted communities.
Forecasting the Air Quality Impacts of Transportation Projects
In Non-Attainment Areas, EPA requires regional and project-level conformity analysis for transportation projects that contribute to traffic emissions. Regional conformity analysis requires forecasting total vehicle emissions using traffic volume forecasts from regional travel demand models and the Motor Vehicle Emission Simulator model developed by EPA (currently MOVES 4) or, in California, a similar Emission Factor model developed by the California Air Resources Board (currently EMFAC2021).72 While regional emission inventories are relatively easy to create, they are
___________________
72 EPA. 2024. “MOVES and Mobile Source Emissions Research.” Updated July 9. https://www.epa.gov/moves; EPA. 2022. “Official Release of EMFAC2021 Motor Vehicle Emission Factor Model for Use in the State of California.” Federal Register, November 15. https://www.federalregister.gov/documents/2022/11/15/2022-24790/official-release-of-emfac2021-motor-vehicle-emission-factor-model-for-use-in-the-state-of-california.
inadequate for evaluating community-scale and neighborhood-scale air quality impacts from transportation projects. Therefore, EPA also requires transportation agencies to conduct project-level conformity analysis for transportation projects that are expected to have significant local air quality impacts, particularly projects expected to produce heavy truck traffic.
Project-level conformity analysis involves forecasting the concentration of traffic emissions near transportation projects to identify local emission hotspots. This evaluation requires an additional modeling step using an air dispersion model, AERMOD.73 AERMOD is used to estimate how future traffic emissions impact local air quality, including if these future emissions would increase ambient concentrations in the surrounding community to levels that exceed the NAAQS.74 For larger projects where air quality is a concern, the National Environmental Policy Act and other state-level environmental regulations may or may not also result in project-level air quality analysis (regardless of an area’s attainment status). However, there is no specific federal requirement to conduct such analysis outside of Non-Attainment Areas. Therefore, neighborhoods experiencing localized air quality issues in a region that otherwise has acceptable air quality can be overlooked.
Adapting Air Quality Models for Equity Analysis
There are several opportunities to adapt methods and models currently used for project-level conformity analysis to evaluate equity concerns related to TRAP exposure caused by transportation projects and plans. Although all these methods involve modeling a proposed project’s impacts on air quality, EPA has developed tools to make this analysis easier for nonexperts, such as local governments and community groups.
Applying Model Analysis to More Project Types, Pollutants, and Areas
Currently limited to large transportation projects in Non-Attainment Areas, project-level air quality analysis using air dispersion modeling can be used for a wider range of transportation projects and outside of Non-Attainment Areas. Even if regional emissions remain unchanged, project-level or location-specific air quality analysis can help identify strategies to mitigate air quality impacts on communities of concern, such as rerouting
___________________
73 The American Meteorological Society and EPA collaborated to develop the AERMOD; it is EPA’s preferred model for air dispersion. https://gaftp.epa.gov/Air/aqmg/SCRAM/models/preferred/aermod/aermod_mfd.pdf.
74 EPA. “Air Quality Dispersion Modeling—Preferred and Recommended Models.” https://www.epa.gov/scram/air-quality-dispersion-modeling-preferred-and-recommended-models. Accessed October 3, 2024.
trucks around—instead of through—neighborhoods surrounding ports.75 In addition, while fine particulate matter is typically the focus of project-level air quality analysis, it is also possible to evaluate a range of other air pollutants including nitrogen dioxide and various hazardous air toxics that are known rise to high concentrations along roadways and that pose significant public health risks.76 Although traffic forecasts are necessary to conduct such project-level analysis, most MPOs already use a travel demand model to forecast the future traffic volumes expected from the projects in their long-range transportation plans or from major infrastructure projects.
The required vehicle emission and air quality models are available from EPA at no cost; however, creating the necessary model inputs, running the models, and interpreting outputs can be a significant burden for smaller transportation agencies or those lacking prior air quality analysis expertise. To address this barrier, EPA has created a reduced-form, simplified, web-based transportation air quality modeling tool called Community-LINE (C-LINE) specifically for local agencies or community groups. Designed for what EPA calls an “initial assessment,” C-LINE can be used to identify at-risk populations near roadways and the impacts of changes in traffic and resulting air pollutions on them.77 C-LINE requires no previous modeling experience.
Applying Project-Level Analysis to Regional Plans
Recent advances in computing capabilities and techniques also make it possible to extend project-level air quality analysis to entire regions, thus enabling evaluation of the local air quality impacts of regional transportation plans.78 Powerfully, evaluating project-level air quality impacts during
___________________
75 Karner, A., D. Eisinger, S. Bai, and D. Niemeier. 2009. “Mitigating Diesel Truck Impacts in Environmental Justice Communities.” Transportation Research Record 2125(1):1–8. https://doi.org/10.3141/2125-01.
76 Karner, A. A., D. S. Eisinger, and D. A. Niemeier. 2010. “Near-Roadway Air Quality: Synthesizing the Findings from Real-World Data.” Environmental Science & Technology 44(14):5334–5344. https://doi.org/10.1021/es100008x.
77 EPA. 2024. “Community-LINE Source Model (C-LINE) to Estimate Roadway Emissions.” Updated June 5. https://www.epa.gov/healthresearch/community-line-source-model-c-line-estimate-roadway-emissions.
78 Cook, R., V. Isakov, J. Touma, W. Benjey, J. Thurman, E. Kinnee, and D. Ensley. 2008. “Resolving Local-Scale Emissions for Modeling Air Quality Near Roadways.” Journal of the Air & Waste Management Association 58(3):451–461. https://doi.org/10.3155/1047-3289.58.3.451; Hatzopoulou, M., J. Hao, and E. Miller. 2011. “Simulating the Impacts of Household Travel on Greenhouse Gas Emissions, Urban Air Quality, and Population Exposure.” Transportation 38(6):871–887; Rowangould, G. 2015. “A New Approach for Evaluating Regional Exposure to Particulate Matter Emissions from Motor Vehicles.” Transportation Research Part D: Transport and Environment 34:307–317. https://doi.org/10.1016/j.trd.2014.11.020.
regional planning can be used to analyze different transportation and land use scenarios and their differences in community exposure to air pollution.79 Conducting this analysis during regional planning also enables the evaluation of emission exposure hotspots and any associated equity concerns before proposed projects move from the long-range transportation plan to the funded program of projects and to the design stage, where there are typically fewer opportunities to address inequitable impacts. As earlier, although the necessary traffic forecasts are generally available and the modeling tools are available from EPA at no cost, the knowledge required to implement such analysis can be a significant barrier. However, larger MPOs that currently perform air quality analysis should have the resources to complete this type of analysis.
Using Air Quality Model Outputs to Forecast Health Outcomes
The output of project-level analysis for a community or region can also be combined with health impact modeling functions and public health data to create metrics that estimate the change in health outcomes attributable to changes in community-scale vehicle emissions exposure.80 EPA’s Environmental Benefits Mapping and Analysis Program (BenMAP) is a free tool that state and local agencies can use to evaluate the benefits of air quality policies by performing spatially detailed health outcomes analysis.81 However, the original BenMAP still requires a spatially detailed air quality
___________________
79 Nadafianshahamabadi, R., M. Tayarani, and G. Rowangould. 2021. “A Closer Look at Urban Development Under the Emergence of Autonomous Vehicles: Traffic, Land Use and Air Quality Impacts.” Journal of Transport Geography 94:103113. https://doi.org/10.1016/j.jtrangeo.2021.103113; Poorfakhraei, A., M. Tayarani, and G. Rowangould. 2017. “Evaluating Health Outcomes from Vehicle Emissions Exposure in the Long Range Regional Transportation Planning Process.” Journal of Transport & Health. https://doi.org/10.1016/j.jth.2017.05.177; Rowangould, D., G. Rowangould, and D. Niemeier. 2018. “Evaluation of the Health Impacts of Rolling Back a Port Clean Trucks Program.” Transportation Research Record. https://trid.trb.org/view/1497247; Tayarani, M., A. Poorfakhraei, R. Nadafianshahamabadi, and G. Rowangould. 2016. “Evaluating Unintended Outcomes of Regional Smart-Growth Strategies: Environmental Justice and Public Health Concerns.” Transportation Research Part D: Transport and Environment 49:280–290. https://doi.org/10.1016/j.trd.2016.10.011; Tayarani, M., R. Nadafianshahamabadi, A. Poorfakhraei, and G. Rowangould. 2018. “Evaluating the Cumulative Impacts of a Long Range Regional Transportation Plan: Particulate Matter Exposure, Greenhouse Gas Emissions, and Transportation System Performance.” Transportation Research Part D: Transport and Environment 63(August):261–275. https://doi.org/10.1016/j.trd.2018.05.014.
80 Poorfakhraei, A., M. Tayarani, and G. Rowangould. 2017. “Evaluating Health Outcomes from Vehicle Emissions Exposure in the Long Range Regional Transportation Planning Process.” Journal of Transport & Health. https://doi.org/10.1016/j.jth.2017.05.177.
81 EPA. 2024. “BenMAP Downloads” Updated May 17. https://www.epa.gov/benmap/benmap-downloads.
forecast as input, which presents that same barrier for smaller agencies as discussed above. In response, EPA has also created a community edition called BenMAP-CE that allows simpler analysis for agencies that are unable to forecast changes in air pollutant concentrations.82
Leveraging Lower-Cost Air Quality Sensors
Because the publicly owned, regulatory air quality monitors required by the Clean Air Act are primarily designed to monitor regional air quality, are few in number, and are expensive, there are significant gaps in data on air quality and TRAP in neighborhoods. New, lower-cost sensors can collect the granular data needed to identify hotspots and monitor progress. A regulatory monitor for one pollutant can cost $50,000 per year to operate, and there may be only a handful of them in a metropolitan area of thousands of square miles. More recently developed sensors, available at lower cost and a smaller, portable size, can be dispersed to more locations to capture additional air quality measurements. Moreover, these new sensors are easier to use, allowing community groups and “citizen scientists” to monitor local air quality. EPA supports the use of these sensors to “help the public learn more about air quality in their communities” and works to advance their capabilities and develops best practices for their use.83 Because this emerging technology is easy for nonscientists to use, these sensors also facilitate the development of crowdsourced air quality data platforms.84
EPA, researchers, monitoring companies, and others are evaluating the reliability of low-cost sensors and their use by community groups.85 In addition, EPA funded 132 community air quality monitoring projects in 2022 as part of a special, one-time Enhanced Air Quality Monitoring
___________________
82 Ben-MAP CE quantifies deaths, emergency department (ED) visits, hospital admissions, heart attacks, respiratory symptoms, asthma attacks, school absences, and lost work days. EPA. 2024. “Environmental Benefits Mapping and Analysis Program—Community Edition (BenMAP-CE).” June 17. https://www.epa.gov/benmap.
83 EPA. 2024. “Air Sensor Toolbox.” Updated May 29. https://www.epa.gov/air-sensor-toolbox.
84 PurpleAir. “Hyper-Local, Real-Time Air Quality Data for Everyone.” https://www2.purpleair.com. Accessed July 24, 2024.
85 Kimbrough, S., S. Krabbe, R. Baldauf, T. Barzyk, M. Brown, S. Brown, C. Croghan, M. Davis, P. Deshmukh, R. Duvall, S. Feinberg, V. Isakov, R. Logan, T. McArthur, and A. Shields. 2019. “The Kansas City Transportation and Local-Scale Air Quality Study (KC-TRAQS): Integration of Low-Cost Sensors and Reference Grade Monitoring in a Complex Metropolitan Area. Part 1: Overview of the Project.” Chemosensors 7(2):26. https://doi.org/10.3390/chemosensors7020026; Commodore, A., S. Wilson, O. Muhammad, E. Svendsen, and J. Pearce. 2017. “Community-Based Participatory Research for the Study of Air Pollution: A Review of Motivations, Approaches, and Outcomes.” Environmental Monitoring and Assessment 189(8):378. https://doi.org/10.1007/s10661-017-6063-7.
Competitive Grant Program. Grant awardees include community-based nonprofit organizations and state, local, and tribal governments for projects using low-cost sensors to help ensure that “overburdened communities have the tools they need to better understand air quality challenges in their neighborhoods.”86
Although lower-cost air quality sensors can create a better picture of TRAP’s impact on neighborhoods, schools, parks, and other places where people are outdoors, the data generated are only useful for establishing baselines and monitoring progress or deterioration. To evaluate project impacts during the transportation planning process still requires the ability to estimate project-driven changes in air quality at the neighborhood or site-specific level and thus TRAP’s impact on health outcomes. Still, a more granular understanding of TRAP-related hotspots and the ability to monitor and evaluate the effectiveness of TRAP-related mitigation strategies are important steps forward and may be particularly valuable for communities experiencing localized air quality issues in a region which otherwise attains NAAQS.
Using Simplified Approaches When Necessary
In cases where the more intensive air quality modeling and data gathering described above is impractical, simplified approaches based on proximity to roadways may be appropriate. The simplest approach is to define a buffer area around high-volume roadways and conduct a demographic analysis on the areas within the buffer. National studies use a 500-meter buffer around roadways with greater than 25,000 annual average daily traffic. However, such methods treat all traffic—and proposed traffic increases—as equally polluting. Currently, cars are becoming cleaner and heavy trucks—the most polluting vehicle and least likely to become electrified in the near term—are becoming more numerous. Proximity analysis methods are under development that can distinguish roadways by mix of traffic and that can model light-duty, medium-duty, and heavy-duty vehicle emissions separately.87
___________________
86 EPA. 2022. “Biden-Harris Administration Announces $53 Million for 132 Community Air Pollution Monitoring Projects Across the Nation.” November 3. https://www.epa.gov/newsreleases/biden-harris-administration-announces-53-million-132-community-air-pollution; EPA. 2024. “Selections for the ARP Enhanced Air Quality Monitoring Competitive Grant.” Updated July 17. https://www.epa.gov/arp/selections-arp-enhanced-air-quality-monitoring-competitive-grant.
87 Antonczak, B., T. Thompson, M. DePaola, and G. Rowangould. 2023. “2020 Near-Roadway Population Census, Traffic Exposure and Equity in the United States.” Transportation Research Part D: Transport and Environment 125:103965. https://doi.org/10.1016/j.trd.2023.103965; Rowngould, G. 2013. “A Census of the US Near-Roadway Population: Public Health and Environmental Justice Considerations.” Transportation Research Part D: Transport and Environment 25:59–67. https://doi.org/10.1016/j.trd.2013.08.003.
An alternative to roadway buffer analysis is to use traffic density, defined as vehicle-miles traveled per unit area such as a census block. Traffic density analysis has advantages over roadway buffer analysis in that it avoids defining buffer thresholds, as it is a continuous metric that can be applied to an entire region. Applications under development can compare traffic by vehicle type and resulting TRAP exposure across geographic units as small as census blocks. When combined with demographic analysis, the traffic density analysis method has advantages for equity analysis. With more precise traffic exposure data, such as where heavy truck traffic uses local roads to access the interstates, the traffic density analysis method’s advantages over the roadway buffer method increase.88
Analyzing Disparate Impacts
For equity purposes, an important follow-up step to any of these methods of reviewing emissions exposure is to calculate whether disadvantaged population groups are impacted differently than the general population. Although required by the 1994 Executive Order on Environmental Justice,89 many metropolitan transportation plans lack detail in this step.90 A test of statistical significance, such as the location quotient, has been federal guidance for decades, but with a few exceptions it is not in widespread use. Location quotients can be used to identify if and where populations of interest are concentrated in areas detrimental to health, and to define whether there is a disproportionately high and adverse effect on a demographic group. A high location quotient points to a significant difference between the proposed project’s impact on the general population and its impact on the overburdened population and could be used as a red flag prompting the transportation agency to examine whether plan alternatives or mitigation strategies are warranted to avoid or reduce further impacts on the TRAP-impacted area.
___________________
88 Antonczak, B., T. Thompson, M. DePaola, and G. Rowangould. 2023. “2020 Near-Roadway Population Census, Traffic Exposure and Equity in the United States.” Transportation Research Part D: Transport and Environment 125:103965. https://doi.org/10.1016/j.trd.2023.103965.
89 Executive Order 12898. February 11, 1994. https://www.archives.gov/files/federal-register/executive-orders/pdf/12898.pdf.
90 TCRP 214 found that few MPOs produce quantitative analyses of project impacts on populations of interest and therefore cannot determine whether these impacts are “disproportionately high and adverse” as required by environmental justice analysis. TCRP 214 recommends location quotients as one method to fill this gap in quantitative analysis (see TCRP Research Report 214: Equity Analysis in Regional Transportation Planning Processes, Volume 2: Research Overview. 2020. Washington, DC: The National Academies Press. https://doi.org/10.17226/25886).
CHAPTER SUMMARY
This chapter presents emerging practices for assessing equity outcomes with innovations in three areas—proximity to destinations, transportation insecurity, and environmental justice—that have promise to significantly advance equity assessment for underserved and disadvantaged populations.
Proximity-to-destinations metrics can be linked to important destinations such as health care facilities, education sites, and employment locations. Proximity-to-destinations indicators are an improvement over transportation network travel-time savings for understanding the opportunities available to communities. This is particularly true when resources and exposures have been inequitably distributed as a result of historic land use and transportation decisions. Ubiquitous data and enhanced computing power make it much more feasible to integrate these metrics into transportation decision making. Proximity or opportunity measures can be defined by time or by space. Both approaches are well established in the research literature and gradually are being adopted in practice. First, cumulative opportunity indicators rely on thresholds to “count” the number of destination types (e.g., jobs or health care facilities) within a travel-time threshold from a specific location via a specific mode (e.g., number of jobs within 45 minutes of a census block group via driving or transit). Second, travel time opportunity indicators provide the travel time to a specified location via a range of modes, such as the travel time from a census block to the nearest hospital or third nearest grocery store via driving or transit. USDOT’s ETC Explorer, the University of Minnesota’s Accessibility Observatory, and other commercial providers have begun the work of making proximity-to-destinations indicators more widely available for an expanding set of destination types. Expanding the applicability and data emerging from these resources should be encouraged; this expansion can hasten the adoption of new equity measures in practice.
The concept of transportation insecurity and the development of tools to measure a person’s transportation insecurity have emerged from research on living with poverty. Definitions may vary slightly, but in general transportation insecurity implies a lack of reliable and safe transportation options. Although the proximity-to-destinations indicators described above can be used to analyze available opportunities to reach destinations, transportation insecurity addresses whether a person’s travel needs have been met. In this sense, transportation insecurity is complementary to proximity measures and can be used to identify less obvious transportation solutions. There is no single source of data in the United States that can be used to systematically identify people for whom the transportation system is failing to meet their needs so severely that they are transportation insecure. University of Michigan researchers have developed the TSI measurement tools. This
approach to assessing transportation insecurity takes its inspiration from decades of work at the national and international levels on food security, which provides models and lessons for operationalizing transportation insecurity on a large scale. The TSI is currently available as a 16-question or 6-question survey module. Both versions have been validated as accurate at identifying people who are transportation (in)secure and for distinguishing levels or gradations of transportation security. Like the 10-question USDA food security survey, the TSI can be used to distinguish the secure from the insecure and to measure levels of insecurity. Work is currently being done to establish the validity of an even shorter, three-question survey tool.
Environmental justice concerns are a subset of equity. The environmental justice impacts of infrastructure development have been attended to in policy and planning processes for quite a while. Most relevant to this committee’s analysis are equity issues evident in disparate impacts, particularly for housing-transportation cost burden, transportation (public) safety, and air quality. With regard to housing-transportation cost burden, to more fully understand transportation’s role in the cost of living, most policymakers and advocates promote metrics that consider the cost of housing and transportation together. There are ways to calculate transportation and housing costs. For example, separate housing and transportation cost burden indicators are included in the USDOT ETC Explorer and are calculated nationwide at the census-tract level. There are issues with data availability to resolve. Both the H+T Index and the Location Affordability Index, measures that could supplement related existing environmental justice screening indicators, deploy peer-reviewed methodologies that rely on aggregate data to model costs over census geographies from a limited set of variables. Although these methods provide comprehensive coverage of housing and transportation using readily available data, the aggregated data can mask variation in household-level characteristics. Housing costs can vary widely across different settings. The geography at which the housing cost burden can be computed will mask these variations at larger zonal scales or geographies. It is important to understand the variability in housing costs in the application of these metrics.
With respect to transportation safety and personal security, reconceiving safety indicators according to the precepts of the safe system approaches and public health tools such as the Safe Systems Pyramid could lead to selecting and deploying indicators—and potentially projects that bring a stronger alignment between public health and road safety. The Pyramid recognizes interventions along the two continuums of individual behavior and population health effects. Both interventions are within the given context of a location or area. Data advances are within reach for more advanced metrics, including linking hospital records to police report data near real-time data from emergency departments across the United States.
Extending these databases to be more accurate for real-time reporting may help in developing effective safety interventions. Crowdsourcing is another way to develop a more comprehensive data set of crashes and safety hazards that include the minor injury, single-person, and near-miss events that are most likely to never make it into police reports. Finally, data needs can be addressed through combining volunteered geographic information with community engagement. Street Story is an example of a community engagement tool developed by the Safe Transportation Research and Education Center at the University of California, Berkeley, that can be used to gather safety data.
Motor vehicle traffic continues to be a large source of local and regional air pollution in the United States. Exposure to TRAP is correlated with adverse health outcomes, including heart disease, lung cancer, and asthma. Advances in mobile source emission and air quality analysis methods are more commonly deployed in regions that fail to attain NAAQS, where more stringent air quality analysis is often required. These advanced techniques could be used to identify environmental justice “hotspots” and evaluate project-level impacts and mitigation strategies as well as deployed in planning stages to analyze transportation and land use scenarios and project alternatives. The adoption of easy-to-use, low-cost air quality sensors can develop more granular data on air quality while empowering local governments and community groups. Project-level air quality analysis using air dispersion modeling can be used for a wider range of transportation projects and outside of Non-Attainment Areas. Even if regional emissions remain unchanged, project-level or location-specific air quality analysis can help identify strategies to mitigate air quality impacts on communities of concern, such as rerouting trucks around—instead of through—neighborhoods surrounding ports.
To produce more equitable transportation infrastructure and services, the metrics derived from the indicators and data improvements discussed in this chapter must be integrated in a meaningful way into transportation decision-making processes, such as regular reporting similar to the Safety PM’s and NAAQS. Achieving equitable outcomes requires that planning, design, and programming decisions are aligned with equity goals, and that these policy goals are implemented at the project level, where the community experiences their benefits and burdens.
