Livable Transit Corridors: Methods, Metrics, and Strategies (2016)

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

109 This Appendix discusses the metrics, methods, and data sources this Handbook uses to mea- sure transit corridor livability. This Appendix focuses on 11 of the 12 metrics used in Step 2 of this Handbook and in the Calculator. These measures and metrics were identified and selected based on the following criteria: • Metrics found in research literature that were theoretically consistent with Transit Corridor Livability Principles, their people and place factors, and the Transit Corridor Livability Goals (described in Step 2). • Metrics that would reasonably reflect the values and needs of people who live, work, and recreate in a corridor (relevance). • Metrics that reflect the needs of a variety of stakeholders and corridor contexts (transfer- ability). • Metrics that are accurate and balanced when considering multiple goals (Haas and Fabish 2013). • Metrics that are appropriate for corridor-level analysis. • Metrics that are useful as performance measures for strategies. • Metrics that are relatively easy for Handbook users to calculate. • Metrics that use readily available data and can be obtained at a low cost to the Handbook users (Haas and Fabish 2013). Table F-1 lists this Handbook’s measures and metrics, the Transit Corridor Livability Prin- ciple each one illustrates, and data sources best used to calculate them. Each metric is categorized according to the concept it is measuring (that is, each metric has a “measure” name). High-Quality Transit, Walking, and Bicycling Opportunities Measure: Corridor Transit Frequency of Service Metric: Aggregate Frequency of Transit Service per Square Mile Calculation Method(s). Because of the complicated process involved in calculating this metric, it is recommended that analysts consult the EPA’s Smart Location Database Version 2.0 User Guide (available at http://www2.epa.gov/sites/production/files/2014-03/documents/sld_ userguide.pdf). Modifications to this metric’s values for the purposes of scenarios analysis should be done using the methods EPA used to calculate this metric. Transit travel time data are typically calculated by MPOs for travel demand modeling forecasts. A P P E N D I X F Metrics, Methods, and Data

110 Livable Transit Corridors: Methods, Metrics, and Strategies Mixed-Income Housing Near Transit Measure: Housing Unaffordability Metric: Percent of Household Income Spent for Housing Calculation Method(s). This metric was modeled by the Department of Housing and Urban Development (HUD) by estimating block-level aggregate income and block-level aggregate rent by apportioning from block group 5-year American Community survey totals, using the proportion of households and the proportion of renter-occupied housing units, respectively. Corridor totals are aggregated from these block-level estimates. The data for this metric comes from U.S. Census Transit Corridor Livability Principles Metric Data Source(s) High-quality transit, walking, and bicycling opportunities Transit employment accessibility EPA’s Smart Locations Data Set (SLD) 2010 SLD ID D5br: Jobs within 45-minute transit commute, distance decay (walk network travel time) weighted Transit frequency of service coverage (aggregate frequency of transit service per sq. mile) SLD SLD ID D4d: Aggregate frequency of transit service (D4d) per square mile Mixed-income housing near transit Housing unaffordability (percent of income spent for housing) HUD’s Housing Affordability Index Data Set (HAI) SLD ID hh_type1_: housing cost as a percent of income for the regional typical household, defined as: Avg. HH size for region, median income for region, average number of commuters per HH for region Income diversity (coefficient of variance of income within corridor) National Historical Geographic Information System (NHGIS), 2010 Census ID B19013: Average of coefficient of variation of block group median household income compared to an average of a corridor’s median Transit-accessible economic opportunities Jobs density (employees/acre) SLD SLD ID D1c: Gross employment density employees (jobs)/acre on unprotected land, 2010 Retail jobs density (retail employees/acre) SLD SLD ID D1c_Ret10: Gross retail employment density employees (jobs)/acre on unprotected land Accessible social and government services Transit corridor ridership balance (RB) Transit agency route/line data Inbound (to CBD) daily boardings/inbound daily alightings Health care opportunities (health care employees/acre) SLD SLD ID D1c8_Hlth10: Gross health care (8-tier) employment density employees (jobs)/acre on unprotected land Vibrant and accessible community, cultural & recreational opportunities Population density (population/acre) SLD SLD ID D1b: Gross population density (people/acre) on unprotected land Access to culture and arts (entertainment employees/acre) SLD SLD ID D1c_Ent10: Gross entertainment employment density employees (jobs)/acre on unprotected land Healthy, Safe and Walkable Transit Corridor Neighborhoods Pedestrian environment (intersection density) SLD SLD ID D3bmm4: Intersection density in terms of intersections having four or more legs per square mile Pedestrian collisions per 100,000 pedestrians Transportation Injury Mapping System (TIMS) 2010 Pedestrian collisions per 100,000 pedestrians Table F-1. Transit corridor livability analysis metrics and associated data sources.

Metrics, Methods, and Data 111 sources, including the 5-year, 2005 to 2010, American Community Survey (ACS) and the 2010 Census (SF1 Form). Block-level data for this metric used in the Calculator and for the research sup- porting this Handbook were obtained from the HUD’s Location Affordability Index dataset. Measure: Racial, Income, Age, and (Dis)Ability Diversity Metric: Corridor/Neighborhood Income Diversity Metric (Coefficient of Variance) Coefficient of Variance (CV) is used to measure income diversity for transit corridors. The CV measures dispersion or how spread out values are from the mean and serves as a standardized method for measuring and comparing income diversity between corridors. CV is defined as the ratio of the standard deviation to the mean for each sample set. As such, the larger CV value, the more dispersion and diversity in corridor incomes. Lower values indicate there is lack of diversity. The data used for this calculation is the 2010 median income for each block group in the United States as recorded in the National Historic Geographical Information System. For the Handbook, the research team segmented out block groups contained within each corridor and performed a separate calculation for each corridor-segmented dataset. The mean and standard deviation for each corridor-segmented dataset were calculated and the ratio of these two numbers produced the CV for that corridor. Transit-Accessible Economic Opportunities Measure: Jobs Density Metric: Corridor/Neighborhood Employees per Acre Calculation Method(s). This metric provides the density of jobs that are accessible within the study corridor or neighborhood. Jobs density can be computed for small areas by dividing population by gross land area. Jobs data can be obtained from the Local Employment Dynam- ics (LED) Database from the U.S. Census. Corridor or neighborhood (block, block group, or census tract) area data can be obtained from the U.S. Census and downloaded from American FactFinder. While these data are easily downloaded from the Internet for geographic units as small as census block groups, they require some effort to aggregate to the corridor level. Measure: Retail Jobs Density Metric: Corridor Retail Employees per Acre Calculation Method(s). This metric provides the density of retail jobs that are accessible within the study corridor or neighborhood. Jobs density can be computed for small areas by dividing population by gross land area. It is calculated in the same manner and uses the same data sources as described above for the Corridor/Neighborhood Employees per Acre metrics. Accessible Social and Government Services Measure: Public Infrastructure and Service Costs Metric: Transit Ridership Balance (A Measure of Transit Corridor Capacity Utilization) The purpose of the Ridership Balance (RB) metric is to provide a measure of transit corridor capacity utilization by gauging the balance (or imbalance) of ridership along a given corridor

112 Livable Transit Corridors: Methods, Metrics, and Strategies and, by extension, present an intra-corridor performance measure of transportation and land use integration. This metric works on the assumption that travel is a derived demand influenced by the accessibility of land use origins and destinations within a transit corridor. The RB metric is a proxy measure of how well a transit corridor investment has been utilized and leveraged by synergistic land use planning actions. In short, the RB metric is a measure of transportation and land use integration, how well the government investments in transportation have been leveraged for land use planning as well as other services (for example, local transit and shuttles). Calculation Method(s). At its core, the RB metric is the ratio of the sum of a corridor’s boardings and alightings, as summed by each station, traveling in a chosen direction, as shown in Equation F-1 below: Equation F-1. RB Metric , ∑ ∑=RB for a chosen direction corridor station boardings corridor station alightings Note that neither CBD stations nor terminus stations are included in the RB metric calcula- tion. Because these stations represent greater, extra-corridor catchment areas and thereby have greater influence on ridership than the intra-corridor stations, they would likely obscure intra- corridor land use/transportation integration performance. Furthermore, the RB measure requires station-level boarding and alighting ridership numbers that are directionally split (for example, inbound, outbound, east, west)—the most important thing to do is to choose a direction for the RB measure for a specific corridor segment and stick to it. For example, if a CBD along the corridor for an inbound measurement is not available, the eastbound ridership data for those stations along that corridor can be used. Where possible, the inbound direction was used, in this instance defined as toward-the-CBD. It is important to note, however, that the RB’s toward-CBD inbound definition does not nec- essarily match the reported format of ridership data, typically assigning an inbound/outbound direction to an entire line which may pass through the CBD rather than having a terminus located there. For example, the LA Metro Rail Gold Line passes through the Los Angeles CBD, from Atlantic station to Sierra Madre Villa (see Figure F-1), rendering a portion of the line’s chosen direction for the RB measure to be actually heading in the outbound direction. Therefore, prior to calculating the RB metric it may be necessary to swap the direction of the boarding and alighting data of some corridors; in other words, reassign the reported outbound-boardings as inbound-boardings and vice-versa (the same applying to the alight- ing data). This is illustrated by the Sierra Madre-Chinatown corridor on the Metro Rail Gold Line (see Figure F-1). The corridor’s reported inbound direction is toward the Sierra Madre Villa station. However, the CBD is situated nearer the opposite end of the corridor (Chinatown Station). Thus, to correctly calculate the corridor’s RB metric, the directionality of this data must first be swapped. The final step in calculating the RB metric is to invert all values greater than “1” (which hap- pens when number of corridor station boardings in the numerator of the equation above is greater than the number of corridor station alightings in the denominator). For all values greater than “1,” simply divide all alightings by all boardings. In doing so, all values will approach “1” as the ridership balance of the corridor is closer to perfect balance.

Metrics, Methods, and Data 113 Measure: Health Care Opportunities Metric: Corridor/Neighborhood Health Care Opportunities (Health Care Jobs per Acre) Calculation Method(s). Health care jobs density can be computed for small areas by dividing the number of health care jobs by gross land area. It is calculated in the same manner and uses the same data sources as described above for the Corridor/Neighborhood Employees per Acre metrics. Vibrant and Accessible Community, Cultural, and Recreational Opportunities Measure: Population Density Metric: Corridor/Neighborhood Population per Acre Calculation Method(s). Population density can be computed for small areas by dividing population by gross land area. Both can come from the U.S. Census and be downloaded from American FactFinder or EPA’s Smart Location Database. While these data are easily downloaded from the Internet for small geographic units (as small as census block groups), they require some effort to aggregate to the corridor level. For existing-conditions analysis, this measure is cal- culated using the ACS data (U.S. Census), using the total population at the block level and then aggregating up to the corridor level. Measure: Corridor Cultural Opportunities Metric: Access to Culture and Arts (Corridor Entertainment Employees per Acre) Calculation Method(s). Entertainment jobs density can be computed for small areas by dividing population by gross land area. It is calculated in the same manner and uses the same Figure F-1. The Gold Line’s reported inbound terminus is Sierra Madre Villa, opposite the toward-CBD direction.

114 Livable Transit Corridors: Methods, Metrics, and Strategies data sources as described above for the Corridor/Neighborhood Employees per Acre metrics. This measure is calculated using the LED Database (U.S. Census), summing the number of Arts, Entertainment, and Recreation (AER) jobs [North American Industry Classification System (NAICS) 71] at the block level and then aggregating up to the corridor level. Healthy, Safe, Walkable Transit Corridor Neighborhoods Measure: Corridor Pedestrian Environment Metric: Corridor/Neighborhood Intersection Density Calculation Method(s). Street connectivity is computed for small areas either in terms of intersection density or percentage of four-or-more-way intersections. Starting with a national dataset of street centerlines, a national database of street intersection locations is produced, including for each intersection feature a count of streets that meet there. Intersections are counted in a GIS program and divided by land area to obtain intersection density. Four-or- more-way intersections are counted and divided by the total number of intersections to obtain the percentage of four-or-more-way intersections in a given area. Measure: Corridor Pedestrian Collisions Rate Metric: Corridor/Neighborhood Pedestrian Collisions per 100,000 Daily Pedestrians Calculation Method(s). Disaggregate collisions data (individual collisions records with latitude/longitude location tags) are currently available for all of California, and the team antic- ipates that similar data will eventually be available for other states. Simple pedestrian collision rates (all casualties, including fatalities and injuries) can be calculated for California corridors by counting the number of pedestrian collisions in a corridor and dividing an estimate of the corridor’s walking population. Equation F-2 provides the calculation formula for this metric. Equation F-2. Daily Pedestrian Collision Rate Formula 100,000 365 = ∗ ∗     PCR PC Pop PS Where, PCR = Daily Pedestrian Collision Rate PC = Total Annual Pedestrian Collisions Pop = Total Population PS = Pedestrian Mode Share percentage from Census Journey to Work data. Data Availability Table F-2 reports on the primary data sources needed by the Handbook users.

Metrics, Methods, and Data 115 Table F-2. Data availability and quality assessment by source. Data Source ApplicableMeasures/Metrics Data Availability Notes U.S. Census/ACS • Population Density (Population/Acre) • Income Diversity (CV of income within corridor) Availability: Excellent Data are easily downloaded from the Internet for small geographic units (as small as census blocks) but requires some effort to aggregate to the corridor level. Data Quality: Excellent Smart Location Database (EPA) • Population Density (Population/Acre) • Employment Opportunities (Corridor Employees/Acre) • Retail Opportunities (Retail Employment/Acre) • Access to Culture & Arts (Corridor Entertainment Employees/Acre) • Corridor Health Care Opportunities (Health Care Employees/Acre) • Pedestrian Environment (Intersections/Acre) • Transit Jobs Accessibility • Transit Frequency of Service Coverage (Aggregate Frequency Of Transit Service per sq. mile) Availability: Very Good Data are easily downloaded from the Internet for small geographic units (as small as census block groups) but requires some effort to aggregate to the corridor level. Data is generally available for U.S. metropolitan areas only. Data Quality: Excellent LED Database (U.S. Census) • Employment Opportunities (Corridor Employees/Acre) • Retail Opportunities (Retail Employment/Acre) • Access to Culture & Arts (Corridor Entertainment Employees/Acre) • Corridor Health Care Opportunities (Health Care Employees/Acre) Availability: Excellent Data are easily downloaded from the Internet for small geographic units (as small as census blocks) but requires some effort to aggregate to the corridor level. Data Quality: Excellent Data for all states available except Massachusetts. TIGER/Line Streets Shapefiles • Pedestrian Environment (Intersections/Acre) Availability: Good Data are easily downloaded from the Internet but requires substantial effort using GIS scripts to count intersections at the corridor level. Data Quality: Very Good Transit Agency Websites • Corridor Line-Haul Mode (“Dummy” Variables for Transit Modes) • Corridor Line-Haul Mode Service Frequencies (Peak Period Average Headways) Availability: Problematic Data are easily found on the Internet, but it is time-consuming to gather data for each agency and route. Data Quality: Very Good (continued on next page)

116 Livable Transit Corridors: Methods, Metrics, and Strategies Data Source ApplicableMeasures/Metrics Data Availability Notes National Resources Inventory Parks and Open Space Inventory Database (USDA) • Corridor Park Coverage (Park Acreage as % of Total Corridor Acreage) • Corridor Park Density (# Corridor Parks/Acre) Availability: Excellent GIS data are easily downloaded from the Internet and purport to provide shape file data on park areas throughout the U.S. Data Quality: Very Good National Transit Database (NTD) • Transit Cost Efficiency (Transit Operating Expense per Person-Miles Traveled) Availability: Good Historical data on transit agency expenses and ridership are easily downloaded from the Internet, but are only available at the agency level; our analysis requires corridor-level data. Data Quality: Problematic Statewide Integrated Traffic Records System (SWITRS) • Pedestrian Collisions Rate (Number of Corridor Pedestrian Collisions per Capita) Availability: Very Good (Outside California: Problematic) California’s SWITRS database provides detailed records of all recorded collisions in the state along with intersection-level geographic identifiers. Unable to find similar statewide dataset for states outside of California. Data Quality: Very Good Table F-2. (Continued).