Dangerous By Design: Methodology
| Dangerous By Design: Full Rankings and Tables | ||
| Table 1: | Most Dangerous Metro Areas (Over 1 Million in Population) | |
| Table 2: | Top 10 Metros with Highest Share of Pedestrian Fatalities | |
| Table 3: | Highest Avg. Fatalities per 100k People Age 65 and Older (by State) | |
| Table 4: | Metros > 1m With Highest Yearly Spending on Pedestrians | |
| Table 5: | State Pedestrian Fatalities & Federal Spending on Walking & Biking | |
| Appendix C: | All 360 Metros with Pedestrian Danger Index Grouped by State | |
| Appendix A: | Methodology | |
<– Back to Dangerous by Design Home
From Appendix A
The National Highway Traffic Safety Administration (NHTSA) collects data on every traffic fatality (pedestrian or otherwise) occurring on U.S. roadways. To determine how many pedestrians were killed in a given year and county, we queried NHTSA’s Fatality Analysis Reporting System (FARS) for pedestrians (and “other pedestrians”) who suffered fatal injuries. We then aggregated the county-level data to the state and Metropolitan Statistical Area (MSA) for 360 metro areas. Dividing this figure by the appropriate population estimate from the U.S. Census Bureau, and multiplying by 100,000 gave us a yearly fatality rate per 100,000 persons. (See the U.S. Census Bureau for definitions of MSA, www.census.gov/population/www/metroareas/metrodef.html)
We utilized the “Pedestrian Danger Index” to allow for a truer comparison of metro areas that takes into account the exposure that pedestrians face in a given metro area. We calculated the Pedestrian Danger Index by dividing the average yearly fatality rate for a metro area by the percentage of commuters walking to work in that metro area. (See the calculation formula below.)
Our exposure measure, the percentage of commuters walking to work, is provided by the U.S. Census Bureau’s 2000 Decennial Census. The ongoing American Community Survey was deemed too unreliable to use (margins of error can be as high as 30% or more). The Census Bureau’s journey-to-work data only captures the trips to work made on foot, excluding trips made on foot to school, to the corner grocery, or to visit friends. Further, respondents to the Census question may select only one mode. So if they walk to the train station or bus stop, and then ride public transit the rest of the way to work, they will likely choose bus or train, rather than walking as their primary mode of transportation to work.
Finally, we calculated spending figures from the Federal Highway Administration’s Fiscal Management Information System (FMIS) – a massive database containing details on all transportation projects that were obligated federal funds. For the purposes of this report, we queried the database for projects with an improvement type related specifically to bicycle and pedestrian programs and facilities (the FMIS database lumps together bicycle and pedestrian projects). The county-level data was then aggregated to the metro area or state level. Dividing this figure by the 2008 population estimate from the U.S. Census Bureau gives us the amount spent on pedestrian projects per capita.
The percentage of federal funds spent on pedestrian projects was determined by dividing this amount derived above for each state by the total federal funds spent in that state (including FHWA funds transferred to FTA for public transportation projects and excluding funds spent on planning and engineering). At the national level, we compared this number to the percent of trips taken by foot, from the 2001 Nationwide Household Travel Survey.
Dangerous by Design Methodology Q & A
Q: How is the Pedestrian Danger Index Calculated?
A: The PDI is calculated by dividing the average annual pedestrian fatality rate per 100,000 residents by the percentage of residents commuting to work on foot. The calculation looks like this:
Average ((2007 total pedestrian fatalities/population)*100,000), (2008 total pedestrian fatalities/population)*100,000))
Percentage of commuters walking to work
Q: What data is used to calculate the Pedestrian Danger Index?
A: We used two data sources for the PDI. First, we queried the National Highway Traffic Safety Administration’s Fatality Analysis Reporting System to determine the number of pedestrians killed in traffic collisions by state and county for the years 2007 and 2008. To determine the number of pedestrians killed in each metropolitan statistical area, we aggregated county-level data to metropolitan areas using Census Bureau definitions.
The second data source we used is the U.S. Census. The Census provides annual population estimates by county, which we utilized to determine the pedestrian fatality rate per 100,000 residents. The Census also estimates the percentage of residents who walk to work. This information is recorded in the journey-to-work question of both the decennial Census and the annual American Community Survey. Unfortunately, the data available through the American Community Survey was deemed too unreliable to use (margins of error can be as high as 30 percent or more), and so we used the most recent decennial Census data, from 2000.
Q: Why are the walk rates so low?
A: Unfortunately, the Census Bureau’s journey-to-work data only captures the trips to work made on foot, excluding trips made on foot to school, to the corner grocery, or to visits friends. Further, respondents to the Census question may select only one mode. So if they walk to the train station or bus stop, and then ride public transit the rest of the way to work, they will likely choose bus or train, rather than walking as their primary mode of transportation to work. Nationally, Americans take about 9 percent of all trips on foot.
Q: How did you calculate the amount of federal funding in the report?
A: We queried the Federal Highway Administration’s Fiscal Management Information System for all obligations, excluding planning and engineering, to determine total obligations within a particular county or state for the fiscal years 2005 through 2008. As with fatalities, county-level data was aggregated up to the appropriate metropolitan area. Bicycle and pedestrian spending was determined by extracting projects classified as a “bicycle or pedestrian facility” improvement type.
Q: Is spending on sidewalk construction as part of a road project included in the spending level?
A: This will depend on the state and how they record the details of a project. If the sidewalk construction component of a project is broken out as a separate element, and the improvement type is specified as “bicycle or pedestrian facility,” then this would be captured by our analysis. If it is lumped together as part of a larger project, and classified as road maintenance or some other improvement type, then it is not included in our analysis.
Transportation for America and the Surface Transportation Policy Partnership have long argued that a better system should be developed, whereby spending of federal funds can be monitored and accurately tracked by the public. The current system, FMIS, is more of an accounting database, with tens of millions of records.
Q: Are state and local spending on bicycle or pedestrian projects included in the analysis?
A: Unfortunately, there is no national data source of this information, and the data available at the state level varies widely in quality and accessibility to the public.
Q: What explains that fact that the top 4 areas in Florida?
A: These metropolitan areas in Florida have high PDI numbers because they have a very high number of pedestrian fatalities and very low rates of people walking. This is likely due to the many high-speed arterial roads in Florida, with few accommodations for pedestrians, such as sidewalks, crosswalks and mid-block crossings.
Q: How does the large population of older adults impact the Pedestrian Danger Index in Florida?
A: Five of the six most dangerous metro areas for walking are in Florida, which is known as a haven for retirees. An analysis of the pedestrian fatality statistics reveals that the portion of older adults dying as pedestrians in Florida is not out of line with the national average. Seventeen percent of pedestrian deaths in Florida in the years studied involved older Americans (70 years and older), the same as the national average of 18 percent. Almost half of the states had rates higher than Florida’s. In fact, over one-quarter of all pedestrian deaths in North Dakota, Hawaii, Vermont, Maine, West Virginia and Massachusetts involved older adults.
