An Institute review of research on crashes involving pedestrians has identified several traffic engineering countermeasures. These can be effective in reducing the approximately 5,000 pedestrian deaths and more than 80,000 injuries that occur each year. About 40 percent of all crashes involving pedestrians occur at intersections.
“Safety officials need to know which options are the most effective for a particular situation,” says Richard Retting, senior transportation engineer at the Institute. “What we’ve done is collect information on what works to reduce the problem, especially at intersections.”
One of the simplest and least costly methods is to change signal light timing to give pedestrians exclusive access to intersections. A study has found that this reduces crashes by about half.
A variation known as early release signal timing gives pedestrians a head start across an intersection before the light changes for vehicle traffic (see Status Report, March 13, 1999; on the web at www.iihs.org). One study found this reduces conflicts between pedestrians and vehicles by about 95 percent.
Another effective approach is to separate vehicles and pedestrians by space - for example, by adding refuge islands, which are raised medians between lanes of opposing traffic. They effectively reduce crashes at wide intersections or where elderly pedestrians cross frequently. A study found that adding such islands reduces pedestrian crash rates by about half at both marked and unmarked crossings. Another study found conflicts reduced by two-thirds.
Roundabouts have been shown to reduce vehicle-to-vehicle crashes when they replace traffic signals and stop signs (see Status Report, July 28, 2001; on the web at www.iihs. org). Two studies also have reported reductions in pedestrian crashes of about 75 percent after conversion to roundabouts.
“Re-engineering an intersection with a signal to a roundabout can be a major undertaking. It can be costly. But it can save money, too, if it means traffic moves more smoothly and there’s no need to add lanes to alleviate congestion,” Retting points out.
Enhancing roadway lighting is an obvious way to improve pedestrian visibility after dark. This is important because more than half of fatal pedestrian crashes occur at night. At crosswalks where the lighting intensity has been increased, two studies found that crashes decreased by more than half.
Other approaches have shown promise but haven’t been as thoroughly evaluated. Advance stop lines, flashing lights in the pavement, and automatic pedestrian detection are examples that merit further study.
Some methods have been ineffective in reducing collisions involving pedestrians. Adding crosswalk markings at locations without traffic signals doesn’t help. One study found that such markings make no difference on narrow, low-volume streets and are associated with higher crash rates on wider roads where more than 12,000 vehicles pass per day. Another study found an increase in crash risk among pedestrians 65 and older when they used marked crosswalks at intersections without traffic signals.
Traffic calming methods have been used with success to reduce vehicle speeds. Such measures include curbs protruding into lanes to make them narrower at crosswalks, speed humps, and islands between lanes (see Status Report, May 2, 1998; on the web at www.iihs. org). But these measures haven’t been shown to reduce pedestrian crash frequencies.
“It’s logical to assume that crash severity would be lessened where traffic calming has slowed vehicle speeds, but this hasn’t been addressed in the research,” Retting says. He adds that “priority among countermeasures should be given to the most effective ones. Traditional approaches without proven effectiveness or approaches that are inappropriate for a given situation will simply waste scarce resources instead of helping to protect pedestrians.”
For a copy of “A review of evidence-based traffic engineering measures designed to reduce pedestrian-motor vehicle crashes” by R.A. Retting et al., see American Journal of Public Health 93:9 (2003).
