We’ve always said that “you measure what you treasure” and the limited, incomplete data about the deaths of people walking suggest that it’s simply not a crisis that our nation cares about. But that doesn’t need to be the end of the story. Here are some specific recommendations to bring things out of the dark ages and into the modern age, while making it clear that reducing these deaths is a top priority for transportation agencies at all levels.
The findings in Dangerous by Design are limited by the data that are publicly available. The best available data lacks clarity, scalability, representation, and is incomplete. The data in DBD is pulled from FARS, which stands for the Fatality Analysis Reporting System. This data is released on a yearly basis containing data from two years prior (meaning, we used 2022 data for the 2024 release of our report). Not only is this data not current, it also fails to break down the number of pedestrian fatalities by race (prior to 2019 for this specific variable), age, income, and other socioeconomic variables. The biggest advantage of FARS data is that it’s geolocational, meaning that we can pinpoint where the most pedestrian-related traffic fatalities are occurring. Any additional context needs to be supplemented with other sources like census data.
Big picture improvements
Overall, crash reports should not be the only form of pedestrian safety data that is widely available at a national level. Inadequate data systems maintained by most states limit quality assurance measures and allow inaccurate data to make it into national datasets. Ideally, the data released by the National Highway Transportation Safety Administration (NHTSA) each year would be current (i.e, including the current year instead of lagging behind by 16-18 months ), and it would include additional variables to contextualize pedestrian fatality data. We wrote more about the limits of federal data reporting here.
Strengthened data sets could also include:
Localized traffic volume data
One of the biggest constraints for pedestrian safety interventions is knowing how many vehicles are on the road. Many pedestrian safety interventions are halted before they get off the ground, with local policymakers citing negative impacts on traffic flow. By normalizing injury rates based on traffic volume, municipalities can better understand which street designs are truly dangerous, as opposed to those that only appear in high-traffic areas.
More information on injuries and injury severity
There is no national dataset on injuries to people walking or rolling that is reliable and comprehensive. Since traffic-related injury data is tracked independently by local police departments and compiled at the state level, all with different recording methods and organizational categories, this data is very difficult to aggregate. Since FARS data can’t track traffic-related injuries at a national scale, the system uses an algorithm to estimate the number of injuries based on the total number of fatalities it has, which, as mentioned earlier, come with its own caveats. Additionally, many traffic-related injuries are not recorded if they are not severe (the definition of “severe” varies from jurisdiction to jurisdiction), and individuals that are treated at emergency rooms or hospitals are often not recorded. This means that individuals that were involved in crashes that waited to get treatment, say, for a concussion or lingering back injury, are not tracked in FARS data either. There is also no dataset that tracks the number of long-term side effects or injuries that people involved in traffic violence experience. Though ER treatment and other hospital data may be able to fill in some of these gaps, this data is neither comprehensive nor publicly available due to privacy concerns.
A 2011 estimate from the International Transportation Forum shows that about 75 percent of pedestrian injuries result from pedestrians being injured while tripping or stumbling on poorly maintained sidewalks and poorly designed streets. In 2020 alone, there were an estimated 54,769 pedestrians injured in traffic crashes. Only about 12 percent of these were fatal injuries, suggesting that injury data is just as important, if not more important, to measure whether street design improvements create less injury risk for all users.
More details about road design
Crash reports provide limited information about where accidents happen and what the conditions and infrastructure look like. They should include crosswalk placement (if any), intersection design, lighting, and visibility measures. This is especially important to track the relative accessibility of streets based on design elements. As of now, pedestrians with different abilities (those who are blind, wheelchair users, those with mobility devices) are lumped together in the vague, overarching group of “vulnerable road users.” To actively address the fatality epidemic among the disability community, these road users need to have their own datasets.
Data on fatalities that occur on private property
Not accounting for deaths or injuries on privately-owned lands (like parking lots) dramatically underrepresents the severity of the pedestrian safety crisis in the U.S. Much of this information exists in proprietary datasets like confidential police reports, hospital, and other health records, making it inaccessible to the public. Putting more detailed data like this behind paywalls and labeling it as proprietary limits the ability of both advocates and practitioners to implement solutions that really work and address the root cause of the problem.
On a macro scale, data collection and analysis should be focused on people walking, not the vehicle. Federal data ignores the motivation of the pedestrian; where and why they were choosing to move the way they were. How we define pedestrians also has an effect on the number of people who are counted in pedestrian death counts. The discrepancies between the state and federal pedestrian fatality and injury data collection systems causes confusion when compiling data. The different definitions used in recording and reporting fatalities versus those used by planners and transportation engineers creates a distorted impression of the relative risk of walking as a mode of transportation. This is more than likely to lead to policies that do not support pedestrian safety investments.