#75 The built environment and pedestrian safety in the Philadelphia region
- Principal Investigator
- Erick Guerra
- Status
- Completed
- Start Date
- July 1, 2017
- End Date
- Jan. 31, 2019
- Project Type
- Research Applied
- Grant Program
- FAST Act - Mobility National (2016 - 2022)
- Grant Cycle
- Mobility21 - University of Pennsylvania
- Visibility
- Public
Abstract
Every year, car collisions claim approximately one and a quarter million lives (World Health Organization 2014). Improvements in vehicle technology and required in-vehicle safety features have driven a decline in U.S. fatalities, but done little to reduce pedestrian traffic fatalities, which account for 14% of the total. In cities with the lowest fatality rates, pedestrians account for an even greater share of fatalities. The share is even higher in dense cities, which tend to have the lowest fatality rates. At the request of newly elected Mayor Jim Kenney, the Philadelphia Streets Department has taken the leading role in sponsoring research and seeking support to develop a strategic framework and toolkit of policies and investments to increase traffic safety. The emphasis is on pedestrians who account for 40% of fatalities. This investigation will support these efforts and seek to create generalizable knowledge that can be applied in a variety of contexts. Matching crash records to spatial measures of the built and social environment, the proposal will examine how the built environment, poverty, exposure, and local enforcement contribute to concentrations of pedestrian injuries and fatalities in Philadelphia using predictive statistical models. In addition to informing policies to reduce pedestrian injuries and fatalities in Philadelphia, the study will focus on how differing units of analysis (road segments, intersections, census tracts, grid cells, etc.) and measures of exposure (pedestrian mode share, traffic volumes, population density, etc.) influence findings. Despite some consistency in findings, there is little consistency in how the existing literature assesses the relationship between the built environment and pedestrian injuries and fatalities.
Description
Every year, car collisions claim approximately one and a quarter million lives (World Health Organization 2014). While this number is projected to continue to increase (Hyder and Bishai 2012), total fatalities in the United States and other wealthy nations have declined steadily over the past two decades (National Highway Traffic Safety Administration 2016). Improvements in vehicle technology and required in-vehicle safety features have driven the decline in fatalities. These improvements have done little to reduce pedestrian traffic fatalities, which have remained stubbornly high at between 4,000 and 5,000 deaths per year and are an increasing share of total fatalities. Since 2011, pedestrians have accounted for approximately 14% of US traffic fatalities. Although technology has failed to improve pedestrian safety, researchers and policy makers have developed a growing body of knowledge about factors that contribute to pedestrian safety. Speeding enforcement (DeAngelo and Hansen 2014; Pilkington and Kinra 2005), higher walking rates (Jacobsen 2003), and denser built environments (Ewing and Dumbaugh 2009; Dumbaugh and Rae 2009) all appear to play important role. Despite being a dense city with a walkable core, Philadelphia has significant progress to make in improving traffic safety. At around 6 fatalities per hundred thousand, Philadelphia’s fatality rate is well below the national average (11 per 100,000), but higher than peer cities like Boston (2.5), Chicago (5), New York (3.5), or Washington, DC (3). Pedestrians account for 40% of these fatalities. In Roosevelt Boulevard, a major arterial in the northern half of the city, Philadelphia has one of the most deadly and dangerous urban roads in America. Other arterial axes, however, also appear to bear a disproportionate number of pedestrian fatalities and serious injuries. Recently elected Mayor Jim Kenney has committed to a policy to work to reduce the number of pedestrian fatalities in the city to zero. While cities like New York, Boston, and Washington, DC, help to offer a model, there are numerous opportunities to deploy geospatial databases to increase local understanding of the important causes of pedestrian fatalities and develop policy interventions—including those enabled through new technologies—to help reduce the number of pedestrian fatalities in Philadelphia and beyond. The Philadelphia Streets Department has taken the leading role in sponsoring research and seeking support to develop a strategic framework and toolkit of policies and investments to increase traffic safety. This proposal will support these efforts and seek to create generalizable knowledge that can be applied in a variety of contexts. APPROACH AND METHODOLOGY This proposal will focus on two important aspects of pedestrian traffic safety in Philadelphia: the role of the built environment and the role of local traffic enforcement in reducing collisions, injuries, and fatalities. Both analyses will rely on a combination of geospatial data and 4 years of crash statistics (2010-2013) from PennDOT for Philadelphia and four surrounding counties in the region. (PennDOT has provided earlier data as well, but assigned spatial attributes to crash data prior to 2010 differently—making the data incompatible.) (1) The role of the built environment Traffic fatalities and serious injuries tend to be lowest in cities with high population densities, narrow streets, and high shares of pedestrians, cyclists, and transit users (Lee, Zegras, and Ben-Joseph 2014; Wier et al. 2009; Morency et al. 2012; Kim, Brunner, and Yamashita 2006; Ewing and Dumbaugh 2009; Dumbaugh and Rae 2009). Surveying the literature, Ewing and Dumbaugh (2009) argue that dense urban environments lower fatality rates by reducing total driving, reducing travel speeds, and increasing the amount of attention that drivers must pay to the road. As fatalities decrease with density, however, pedestrian fatalities remain fairly flat and the proportion of pedestrian fatalities increases. Figure 1 plots this relationship across several large US cities. Although fatality rates are lower when accounting for higher exposure due to more walking, the burden of harm shifts toward a population that is both more vulnerable and less responsible for generating fatalities. Fatalities also tend to increase in places with lower income households, more children, and more elderly residents (Wier et al. 2009; Aguero-Valverde and Jovanis 2006; Ewing, Schieber, and Zegeer 2003; LaScala and Grunewald 2004). In Philadelphia, high rates of pedestrian fatalities appear to occur in neighborhoods with high walking rates, high concentrations of poverty, and a density of high-speed arterial roads like the notorious Roosevelt Boulevard. Figure 1. The relationship between population density and traffic fatalities across large US cities Matching crash records to spatial measures of the built and social environment, this proposal will examine how the built environment, poverty, and exposure contribute to concentrations of pedestrian injuries and fatalities in Philadelphia using predictive statistical models. In addition to informing policies to reduce pedestrian injuries and fatalities in Philadelphia, the study will focus on how differing units of analysis (road segments, intersections, census tracts, grid cells, etc.) and measures of exposure (pedestrian mode share, traffic volumes, population density, etc.) influence findings. Despite some consistency in findings, these is little consistency in how the existing literature assesses the relationship between the built environment and pedestrian injuries and fatalities. (2) The role of local enforcement In addition to city design, a lack of enforcement may play a role in Philadelphia’s disproportionately high fatality rates. DeAngelo and Hansen (2014) looked at the effects of State Trooper layoffs on traffic safety in Oregon and found a statistically and economical significant relationship. They estimated that an additional $309,000 in spending on State Troopers saved an estimated one life per year. In a review of studies on the effectives of speeding cameras, Pilkington and Kinra (2005), found cameras reduced collisions between 5% and 69%, injuries between 12% and 65%, and fatalities between 17% and 71%. The Philadelphia police, due to other important public safety priorities, rarely enforces traffic violations and current state law prevents the deployment of speeding cameras in the city. The city, however, provides something of a natural experiment on enforcement through the University of Pennsylvania’s special policing zone. The University of Pennsylvania employs approximately 100 full-time publicly sworn officers who supplement the existing public police force in the immediate vicinity of Penn’s campus. Due to University concerns over student injuries, furthermore, these officers may place a greater emphasis on traffic safety—a supposition that will be examined through interviews and an analysis of police citations. If enforcement works, presumably a street on one side of the border will tend to have fewer collisions over time than the same street on the other side of the border. MacDonald et al. (2012) apply a similar approach to estimate the effect of additional policing on crime reduction.
Timeline
This ongoing project has moved from T-Set to the Mobility21 University Transportation Center. Literature review and synthesis (May 2016 - August 2017) Review and summarize existing studies of the relationship between the built environment Data processing (May 2016 - June 2017) Clean crash data and match with existing demographic and built environment measures including the Census, local tax assessor parcel data, the Bureau of Labor Statistics Zip Business Patterns data, and street shapefiles. Time-series Analysis (June 2017 - October 2017) Examination of whether increased population density leads to safer streets within neighborhoods. Cross-sectional analyses: (August 2017 - October 2018) Conduct analysis of relationship between measures of built environment and pedestrian injuries and fatalities using 3-to-4 units of analysis (segments, census tracts, blocks, and grid cells).
Strategic Description / RD&T
Deployment Plan
Findings will be presented at the 2017 Association of Collegiate Schools of Planning in Denver, CO, the Transportation Research Board, and other conferences. I will continue to work with local partners (City of Philadelphia and Delaware Valley Regional Planning Commission) to look for policy relevant findings in the near term. Philadelphia Crash Map to summarize data and findings: http://crashphilly.erickguerra.net/
Expected Outcomes/Impacts
Three peer-reviewed publications. Conference presentations. Influence on the discussion of how the built environment affects traffic safety, particularly for pedestrians.
Expected Outputs
TRID
Individuals Involved
| Name | Affiliation | Role | Position | |
|---|---|---|---|---|
| xiaoxiad@design.upenn.edu | Dong, Xiaoxia | University of Pennsylvania | Other | Student - PhD |
| ilangold@sas.upenn.edu | Gold, Ilan | ilangold@sas.upenn.edu | Other | Student - Undergrad |
| erickg@upenn.edu | Guerra, Erick | University of Pennsylvania | PI | Faculty - Untenured, Tenure Track |
| michelleckondo@fs.fed.us | Kondo, Michelle | USDA Forest Service | Other | Other |
| cmorrison@prev.org | Morrison, Chris | University of Pennsylvania | Other | Faculty - Research/Systems |
Budget
Amount of UTC Funds Awarded
$50000.00Total Project Budget (from all funding sources)
$Documents
Match Sources
No match sources!
Partners
No partners!