DATE: November 5, 2020
TO: Boston Region Metropolitan Planning Organization
FROM: Betsy Harvey
RE: Disparate Impact Metrics Analysis Study
This memo describes the second and final phase of the Boston Region Metropolitan Planning Organization’s (MPO) effort to develop a Disparate Impact and Disproportionate Burden (DI/DB) Policy for the MPO’s Long-Range Transportation Plan (LRTP). The policy will be used to evaluate, in the aggregate, projects funded in the LRTP Recommended Plan for potential future disparate impacts and disproportionate burdens on minority and low-income populations, respectively. The DI/DB Policy responds to Federal Highway Administration (FHWA) and Federal Transit Administration (FTA) requirements that recipients of federal funding identify and address potential future disparate impacts and disproportionate burdens.
Phase Two was undertaken through the MPO’s Disparate Impact Metrics Analysis study. The goal of the study was to revise the thresholds that are used in the draft DI/DB Policy to identify disparate impacts and disproportionate burdens. This included reconvening the MPO’s DI/DB Policy stakeholder working group that was first convened during Phase One. This memo describes a literature review conducted to help revise the thresholds, the outcomes of the working group, staff’s recommendations for the final DI/DB Policy, and summary of the study results. The attached technical appendix gives more detail about the study metrics and thresholds.
The projects that are analyzed with the DI/DB Policy are those included in the LRTP Recommended Plan and that change the capacity of the transportation network. The Recommended Plan consists of regionally significant projects within the Boston MPO region that will be financed with federal funds. Regionally significant projects are defined as
Projects that do not change the capacity of the transportation network, including those funded through the MPO’s investment programs, are analyzed in the Transportation Improvement Program.
The MPO’s LRTP DI/DB Policy responds to two federal mandates: Title VI of the Civil Rights Act of 1964 and Executive Order 12898, Federal Actions to Address Environmental Justice in Minority Populations and Low-income Populations, known as the Environmental Justice Executive Order (EJ EO).
Title VI prohibits intentional discrimination (disparate treatment) and unintentional discrimination (disparate impact) based on race, color, or national origin. Disparate treatment refers to actions that result in circumstances where similarly situated persons are intentionally treated differently because of their race, color, or national origin. Disparate impacts refer to a facially neutral policy or practice that disproportionately affects members of a group identified by their race, color, or national origin, where a funding recipient’s policy or practice lacks a substantial legitimate justification, and where there exists one or more alternatives that would serve the same legitimate objectives but with a less disproportionate effect.
Federal agencies enforce disparate impact compliance through Title VI, implementing regulations with which recipients of an agency’s funding must comply. As recipients of funding from FTA and FHWA, MPOs must comply with both of these agencies’ Title VI disparate impact regulations.
The EJ EO was issued in 1994 and made achieving EJ a mission of the executive branch. The EJ EO directs federal agencies to identify and address disproportionately high and adverse environmental and human health effects of their activities, policies, and practices on minority populations and low-income populations. This obligation is passed to recipients of federal funding through executive branch agencies, including FTA and FHWA. For low-income populations, disproportionately high and adverse effects are referred to as disproportionate burdens.
FTA’s most recent Title VI circular was issued in 2012. It requires MPOs to identify and address disparate impacts on minority populations that may result from projects programmed with state or federal funds in the aggregate. In 2015, FHWA issued combined Title VI and EJ guidance that similarly states that MPOs must assess the impacts on minority and low-income populations that may result from its activities. The DI/DB Policy addresses both FTA’s and FHWA’s requirements. These documents set guidelines for the following analyses:
In 2017, the Boston region MPO committed to developing a DI/DB Policy to use in subsequent LRTPs that would
These changes would not only align the DI/DB Policy with federal guidance but would also ensure that the MPO allocates resources to address meaningful impacts. In addition, staff saw an opportunity to engage members of the public in the development of a new policy to ensure the policy reflected their interests.
MPO staff pursued a two-phased effort to develop the DI/DB Policy. The first phase took place in federal fiscal years (FFY) 2018 and 2019; the second phase took place in FFY 2020. For the first phase, during the spring and summer of 2018 staff began a public engagement process to solicit stakeholder and public input. Staff presented the results of this process to the MPO and solicited members’ input on October 4, 2018.
Staff then synthesized public input to update the DI/DB analysis methodology, revise the metrics that would be analyzed, and quantify uncertainty in the travel demand modeling process. Staff developed a draft DI/DB Policy that incorporated all of this work. It was presented to the MPO on May 2, 2019, and the MPO approved its use in Destination 2040, the current LRTP. (Phase One is described in detail in the memo Development of the DI/DB Policy for the LRTP: Phase One.)
The DI/DB Policy assesses the impacts of several metrics on four population groups—minority and low-income populations (protected populations) and nonminority and non-low-income populations (non-protected populations). During Phase One, MPO staff updated the metrics used in Destination 2040, which were used in this study to help develop the DI/DB Policy’s thresholds.
The draft DI/DB Policy established the following guidelines to determine whether there would be disparate impacts and disproportionate burdens:
[A]ny impact that is projected to adversely affect the protected population more than the non-protected population, and where the MPO can be confident that this is not due to model uncertainty, would indicate a potential future disparate impact or disproportionate burden.
Through this study, staff further refined this approach, proposing the use of three thresholds to identify disparate impacts and disproportionate burdens: the baseline uncertainty threshold, the practical impact threshold, and the disproportionality threshold. This approach clarifies how the MPO identifies disparate impacts and disproportionate burdens and aligns the policy with federal guidance as described in the Phase One memo. The DI/DB Policy must demonstrate that the impact is caused by the LRTP projects, that the impact is significant, and that it disproportionately affects the protected population compared to the non-protected population. The three thresholds address each of those requirements, and each metric must pass all three thresholds for there to be a finding of a disparate impact or disproportionate burden.
The first threshold is the baseline uncertainty threshold. It determines whether the predicted impact for each population group would likely occur or whether it would likely be due to the inherent uncertainty in the modeling process that is used to identify impacts.
The second threshold is the practical impact threshold. It determines whether the impact would be practically significant. (An impact that is practically significant is one that would have a demonstratable effect on people’s quality of life. For example, an increase in carbon monoxide emissions that affects health outcomes.) The change between the no-build and build scenarios for each population group is compared to the practical impact threshold.
The third threshold is the disproportionality threshold. It determines whether the impact would adversely affect the protected population more than the non-protected population.
In Phase One, staff identified the uncertainty associated with each of the 10 metrics that are used in the baseline uncertainty threshold. The purpose of Phase Two was to develop thresholds for the practical impact and disproportionality thresholds. In Phase One, stakeholders stated that they did not want the policy to allow any difference between the impacts to the protected and non-protected populations, unless there was evidence that a certain level of change rose to the level of being significant in terms of its effects on people’s quality of life. Staff conducted a literature review to answer that question.
Identifying practical impact thresholds was a unique challenge because of the character of the distribution of the impacts analyzed in the DI/DB Policy. Because of the large geographic area covered by the analysis, transportation analysis zone (TAZ) values range, for example, from very low to very high travel times—yet the threshold is applied to the regional average. Using an absolute value (such as minutes of travel time) as the threshold would be useful for one part of the distribution (such as short travel times), but less useful for other parts. For that reason, staff sought to develop thresholds that were expressed as percentage changes. (See the Appendix for more details regarding analyses of the distribution of values across TAZs.)
The goal for setting a practical impact threshold for travel time was to determine the magnitude of change in travel time that would be significant. (Because congested VMT can be understood as the amount of extra time a driver spends in traffic compared to an entire free flow trip, this metric was grouped with the travel time metrics.)
There is a large body of travel time literature on commute trips. Studies have shown that longer commutes lead to declines in mental and physical health, including headaches, high blood pressure, poor sleep, and high stress levels, among other symptoms. While every additional minute of commuting time affects well-being, that minute is a larger percentage of travel time for those with shorter commutes than it is for those with longer commutes. Therefore, a percentage change would be a more accurate way to express a threshold.2, Staff did not find any research, however, that suggested such a cutoff.
Staff also explored whether the value of travel time (VTT) could provide useful information for setting the threshold. Because cost is a variable in the travel demand model, travel time can be converted to monetary cost. VTT can be defined as the cost of time spent on traveling. It is most commonly expressed as a percentage of hourly wages. There are many variables that affect the VTT for a given traveler. These include
The model can produce metrics that break out travel times for some of these variables, such as mode, trip purpose, and congestion level, but for other variables, such as traveler preference, it would likely be impractical. Because of the number of factors that contribute to VTT, incorporating all of them into one metric would be impractical for the purposes of this analysis. Setting one value for an entire population would only incorporate a few of these variables and would exclude important differences in the variables that affect VTT. In addition, staff were unable to find any literature that suggested when a change in travel time is significant. Because of the many variables that go into measuring VTT, it did not prove to be a useful avenue through which to develop thresholds for travel time metrics.
Breathing air with high levels of CO reduces the amount of oxygen available to critical organs, such as the heart and the brain. One major source of CO is vehicles. Because of the health effects of CO, especially for sensitive populations, it is prudent to think of air quality as meeting a certain standard developed to reflect health impacts. Staff explored using national and state CO standards as the practical impact threshold.
The MPO measures CO emissions in kilograms per square mile, which is based on the vehicle types, volumes, and trips made within each TAZ. To identify levels that are damaging to health, staff reviewed emission targets set by state and national agencies. Staff focused on primary targets, which provide public health protection, including for sensitive populations (such as people with asthma, children, and the elderly). The Environmental Protection Agency’s (EPA) primary targets are 30 micrograms per meter cubed (mcg/m3) (90 parts per million [PPM]) for one hour and 10 mcg/m3 (9 ppm) for eight hours.4 The Massachusetts Environmental Protection Agency (MEPA) standards are identical to EPA standards.5
Staff found several roadblocks that limited the usefulness of these standards. The standards pertain to data collected from point locations, whereas the impacts generated by the model reflect emissions in all MPO TAZs. In addition, the EPA and MEPA reports CO emissions standards as volumes (parts per million or microgram per meter), which cannot be meaningfully converted to kilograms. Because of these differences, EPA (and MEPA) thresholds for unhealthy levels of CO were not useable for the purposes of the DI/DB Policy.
The accessibility metrics measure access to land uses is critical to residents’ quality of life. While much research analyzes access to these opportunities, as well as differences in access between protected and non-protected populations, nothing indicates how large a reduction in access would be harmful, perhaps suggesting that any decrease would be harmful.6
Staff also explored how other MPOs analyze access to various opportunities, since these types of metrics are common to many MPOs. Staff found that there is no standard across MPOs, and many analyses are qualitative. For example
Staff found that MPOs often qualitatively assess accessibility for protected populations, such as comparing EJ results to total for region without stating whether there’s a DI or DB.7 The San Diego Council of Governments uses a 20 percent threshold to identify disparate impacts and disproportionate burdens—for accessibility metrics as well as all of its metrics—but it is not related to the magnitude of actual impact on population groups.8 The Metropolitan Transportation Commission (the San Francisco area MPO) uses statistical significance.9 In light of this research, staff found there was no indication that suggested that any decrease in accessibility would not be harmful.
As part of this study, staff reconvened the same stakeholder working group that was previously gathered during Phase One. The purposes of the meeting were to (1) gather feedback from stakeholders on staff’s threshold recommendations; (2) illustrate the benefits and drawbacks of using different thresholds through the DI/DB threshold application; and (3) show the role that the policy plays within the MPO’s broader approach to equity.
The meeting took place on August 25, 2020. Stakeholders included Boston region MPO members, representatives from several advocacy groups, and representatives of human service agencies, for a total of 12 participants. The meeting was also open to members of the public; there were 29 non-CTPS staff attendees, including stakeholders. Although the first three stakeholder working group meetings were convened in person, this meeting was conducted via the Zoom online meeting platform because the COVID-19 pandemic prohibited in-person meetings. The first half of the meeting was dedicated to reminding attendees the role of the DI/DB Policy and showing them the proposed revisions to the draft policy, and the second half consisted of guided discussions in small and large groups to gather reactions from stakeholders.
Stakeholders were overall supportive of the proposed changes. Stakeholders provided the following comments:
Stakeholders also provided the following recommendations for future MPO work:
Staff developed the following threshold recommendations for the DI/DB Policy, which were proposed to the stakeholder working group. They are incorporated into the final DI/DB Policy.
In the fall of 2020, staff will seek the MPO’s endorsement of a final DI/DB Policy, which will incorporate the revisions described in this memo. Prior to the next LRTP, staff will also update existing metrics and add new ones that are analyzed for disparate impacts and disproportionate burdens. The MPO will use both the metrics and the final DI/DB Policy in the next LRTP.
The Boston Region Metropolitan Planning Organization (MPO) operates its programs, services, and activities in compliance with federal nondiscrimination laws including Title VI of the Civil Rights Act of 1964 (Title VI), the Civil Rights Restoration Act of 1987, and related statutes and regulations. Title VI prohibits discrimination in federally assisted programs and requires that no person in the United States of America shall, on the grounds of race, color, or national origin (including limited English proficiency), be excluded from participation in, denied the benefits of, or be otherwise subjected to discrimination under any program or activity that receives federal assistance. Related federal nondiscrimination laws administered by the Federal Highway Administration, Federal Transit Administration, or both, prohibit discrimination on the basis of age, sex, and disability. The Boston Region MPO considers these protected populations in its Title VI Programs, consistent with federal interpretation and administration. In addition, the Boston Region MPO provides meaningful access to its programs, services, and activities to individuals with limited English proficiency, in compliance with U.S. Department of Transportation policy and guidance on federal Executive Order 13166.
The Boston Region MPO also complies with the Massachusetts Public Accommodation Law, M.G.L. c 272 sections 92a, 98, 98a, which prohibits making any distinction, discrimination, or restriction in admission to, or treatment in a place of public accommodation based on race, color, religious creed, national origin, sex, sexual orientation, disability, or ancestry. Likewise, the Boston Region MPO complies with the Governor's Executive Order 526, section 4, which requires that all programs, activities, and services provided, performed, licensed, chartered, funded, regulated, or contracted for by the state shall be conducted without unlawful discrimination based on race, color, age, gender, ethnicity, sexual orientation, gender identity or expression, religion, creed, ancestry, national origin, disability, veteran's status (including Vietnam-era veterans), or background.
A complaint form and additional information can be obtained by contacting the MPO or at http://www.bostonmpo.org/mpo_non_discrimination. To request this information in a different language or in an accessible format, please contact
Title VI Specialist
1 The minority population include people who identify as Black/African American; Asian; Hawaiian or other Pacific Islander; Alaskan Native or American Indian; and/or Hispanic. People are considered low income if they live in a family whose annual income is less than or equal to 200 percent of the national poverty level.
2 Annette Schaefer, “Commuting Takes Its Toll,” Scientific American, October 1, 2005, https://www.scientificamerican.com/article/commuting-takes-its-toll/Office for National Statistics, “Commuting and Personal Well-being, 2014,” The National Archives, February 12, 2014, https://webarchive.nationalarchives.gov.uk/20160105231823/http://www.ons.gov.uk/ons/rel/wellbeing/measuring-national-well-being/commuting-and-personal-well-being--2014/art-commuting-and-personal-well-being.html#tab-abstract.
3 Victoria Transport Policy Institute, “Transportation and Travel Time Costs,” in Transportation Cost and Benefit Analysis: Techniques, Estimates, and Implications, March 20, 2020, https://www.vtpi.org/tca/tca0502.pdf.
4 United States Environmental Protected Agency, NAAQS Table, last updated December 20, 2016, https://www.epa.gov/criteria-air-pollutants/naaqs-table.
5 Ambient Air Quality Standards for the Commonwealth of Massachusetts, 310 C.M.R. 6.00 (2019), https://www.mass.gov/doc/310-cmr-600-ambient-air-quality-standards/download.
6 Joseph Galaskiewicz, Kathryn Freeman Anderson, and Kendra Thompson-Dyck, “Minority-White Income Inequality across Metropolitan Areas: The Role of Racial/Ethnic Residential Segregation and Transportation Networks,” Journal of Urban Affairs 0, no. 0 (October 31, 2019): 1–24.https://doi.org/10.1080/07352166.2019.1660581; Armin Jeddi Yeganeh, Ralph Hall, Annie Pearce, and Steve Hankey, “A Social Equity Analysis of the U.S. Public Transportation System Based on Job Accessibility,” Journal of Transport and Land Use 11, no. 1 (November 13, 2018), https://doi.org/10.5198/jtlu.2018.1370.
7 Southern California Council of Governments, Regional Transportation Plan: 2012-2035: Environmental Justice Appendix, April 2012, http://rtpscs.scag.ca.gov/Documents/2012/final/SR/2012fRTP_EnvironmentalJustice.pdf.
8 San Diego Association of Governments, San Diego Forward: The 2019 Federal Regional Transportation Plan, October 2019, https://sdforward.com/docs/default-source/2019federalrtp/draftfinal/app-h---social-equity-engagement-and-analysis.pdf?sfvrsn=247ff65_2.
9 Metropolitan Transportation Commission, Plan Bay Area 2040: Final Equity Analysis Report, July 2017, http://2040.planbayarea.org/sites/default/files/2017-07/Equity_Report_PBA%202040%20_7-2017.pdf.
DISPARATE IMPACT METRICS STUDY APPENDIX