University of Minnesota Driven to Discover
U of MNUniversity of Minnesota
Center for Transportation Studies

More signage may not make crosswalks safer, speaker says

Presentation by Thomas J. Smith, Department of Kinesiology, University of Minnesota
Humphrey Institute of Public Affairs—Regional Planning and Policy Seminar
November 12, 2009

The increase in the number of walkers and bikers in urban areas offers a challenge for planners trying to prevent accidents. Many cities have installed roadway stripes, yellow warning signs, and even flashing lights (active warnings) at uncontrolled pedestrian crosswalks (crosswalks located mid-block or at unsignalized intersections) to signal drivers that pedestrians might be present.

How well do those measures work? A recent study led by Thomas J. Smith of the University’s Department of Kinesiology found that in many cases, drivers do not slow down for marked crosswalks—even those with active warnings.

“A pedestrian may or may not be present in the crosswalk,” Smith said at a Regional Planning and Policy seminar on Nov. 12 at the Humphrey Institute of Public Affairs. “Drivers, if they can’t predict what is coming, are going to pay more attention to whether or not a pedestrian is in proximity to the crosswalk than to the warning sign itself.

“So, you could argue that an active warning is not always a useful improvement at an uncontrolled crosswalk.” he said.

The study found that pedestrians are safer at marked crossings where they can activate warning lights when they cross.

The findings contradict a trend towards more heavily marking uncontrolled pedestrian crosswalks. But they add to a growing body of literature that has found similar results.

Previous research

Pedestrian crashes at uncontrolled crosswalks account for about a quarter of all pedestrian–vehicle crashes, Smith said. Although pedestrian fatalities have dropped in the last 20 years, from 7,000 to about 5,000 annually, they still make up 10 percent of total traffic fatalities.

Smith and his team found previous research on the subject was mixed. Five studies showed that in most cases marked crosswalks were less safe than unmarked crosswalks for pedestrians. In three studies, controlling for pedestrian usage, pedestrian crash levels at marked crosswalks were higher compared with unmarked crosswalks. In one, removal of crosswalk markings resulted in a 61 percent decline in pedestrian crashes.

“These findings seem counterintuitive,” Smith said. But, he added, “We should base crosswalk decisions on empirical evidence, not ad hoc assumptions.”

Metro-wide study

For this project, Smith and his team observed 18 uncontrolled pedestrian crosswalk sites in the Twin Cities metro area from July to November 2007. Nine were sites with passive warnings, six were sites with pedestrian-activated active warnings, and three were sites with continuously active warnings. Video cameras captured more than 7,000 vehicle–crosswalk interactions and some 600 vehicle–pedestrian interactions. Speed limits at the sites were 25, 30, and 35 mph.

The team later analyzed the data to look at changes in both vehicle velocities and accelerations at the crosswalks.

The result: Relative to crosswalk sites with passive warnings, sites with continuously active warnings showed no difference in average vehicle velocities as vehicles approached the crosswalk.

“The only sites where we saw significantly lower average vehicle velocities are those with pedestrian-activated warnings,” Smith said.

The study concluded that the meaning of warnings at uncontrolled pedestrian crosswalks is ambiguous, because drivers pay more attention to the presence or absence of pedestrians at the crosswalk rather than to the type of warning. And the results reinforced previous research that questions the usefulness of active versus passive warnings at uncontrolled crosswalks.

Smith said further analysis is needed, but his study offered two key recommendations: (1) Cities should consider replacing current passive or continuously active warnings at uncontrolled crosswalks with those that are pedestrian-activated, and (2) Cities should consider technology that automatically detects a pedestrian or bicyclist crossing at a crosswalk, without forcing them to stop and push a button.

“There is a cost concern,” Smith said, but he added, “Any advance warning of a biker or pedestrian approaching a crosswalk is likely to have a safety benefit.”

For more information on this study, visit the ITS Institute research page.