Evaluating pedestrian and bicyclist risk in Minnesota roundabouts

Once rare in the United States, roundabouts are becoming more common in Minnesota and across the country. Although research has shown that roundabouts can successfully ease congestion and reduce serious crashes, there are concerns about roundabout accessibility and safety for pedestrians and bicyclists.

In a study funded by the Minnesota Department of Transportation, researchers from the Minnesota Traffic Observatory (MTO) examined the experience of pedestrians and bicyclists at two roundabouts in the Twin Cities. Led by MTO director John Hourdos, the team used video surveillance equipment to collect data on driver and pedestrian behavior at each site.

The researchers deployed the video equipment at a two-lane roundabout in suburban Richfield and a one-lane roundabout in a residential area of Minneapolis. Between the two locations, the team captured video of more than 6,900 pedestrian crossings and 7,500 bicycle crossings.

There are approximately 2,000 roundabouts in the U.S., compared to 20,000 in France, 15,000 in Australia, and 10,000 in the U.K.

Among the thousands of crossing events captured by the study, there were only three cases that could marginally be termed close calls. However, study findings do highlight the existence of friction between pedestrians and drivers at roundabout crossings.

One concern is that many drivers fail to yield to pedestrians and bicyclists, even though Minnesota law requires them to do so. In fact, findings show that drivers at the Richfield roundabout yield only about 45 percent of the time. The yielding rate in Minneapolis was higher, averaging about 83 percent.

The research team identified several factors that influence drivers’ yielding behavior. Study results indicate the following trends:

• Drivers are more likely to yield to pedestrians or bicyclists beginning their crossing in the center island.
• Vehicles exiting the roundabout are less likely to yield than those entering it.
• Drivers are more likely to yield to larger groups.
• Vehicles entering the roundabout at the immediate upstream entrance are more likely to yield than those coming from other entrances.
• Drivers are less likely to yield if they encounter another vehicle merging into the roundabout immediately before the exit where the pedestrian is trying to cross.
• Yielding probability decreases with more vehicles present in the roundabout.

The team also examined the delays experienced by pedestrians waiting to cross at roundabouts, and results indicate that their wait times are actually shorter than at signalized intersections. For example, a signalized intersection with daily traffic comparable to the Richfield roundabout has an average pedestrian delay of 30 seconds. The average delay at Richfield was nine seconds. At Minneapolis, the wait was even shorter—less than two seconds. However, the researchers suggest that the non-yielding behavior of drivers at roundabouts may intensify the experience of delay, making it seem longer than it actually is.

In addition, drivers’ failure to yield creates a significant safety risk, particularly for pedestrians who are visually impaired. Although the researchers did not observe any visually impaired pedestrians in this study, the observed yielding rates demonstrate that such pedestrians cannot assume drivers see them or are willing to stop.

Reprinted with permission from the November 2012 issue of Catalyst, a publication from the Center for Transportation Studies.