In recent discussions and cussing here at Better Burque about safety and bike infrastructure around town, the topic of roundabouts and traffic circles, naturally, has come up. Are these treatments safer than red lights and two/four-way stops?

Keeping in mind that large roundabouts (think Yale just north of Lomas in the UNM campus area) are quite different than traffic circles (think Silver at Cornell), “conventional wisdom,” such as that expressed in this National Highways Traffic Safety Administration newsletter, considers both roundabout and traffic circles more safe than other alternatives.

But a BB reader recently passed along this study from the October 2013 issue of the journal Injury Prevention that found higher rates of bike injuries at smaller traffic circles.

Traffic circles were more hazardous than all other intersection types (traffic lights, two-way stops, four-way stops and uncontrolled intersections). All traffic circles were in Vancouver, at intersections of two local streets (intersections that were otherwise found to be very safe). In Vancouver, local street traffic circles are small (approximately 6–8 m in diameter; figure 4) unlike major street roundabouts common in Europe. Comparisons are difficult because of the size difference, but other studies have shown that large roundabouts (∼30 m in diameter) reduce injury risk for motor vehicle occupants and this has fostered interest in this intersection design.18 19 Despite the reduced risk for motor vehicles at roundabouts, studies of cyclists have found increased risks.20–22 Brüde and Larsson23 showed that roundabouts with radii greater than 10 m were safer than smaller circles. Daniels et al21 and Schoon and Van Minnen19 showed reductions in roundabout risks for cyclists when there were lanes physically separated from motor vehicles. The increased risk to cyclists associated with traffic circles and roundabouts could relate to the large number of associated ‘conflict points’.24 In our study, the two main types of crashes at traffic circles were with motor vehicles (n=8), because the cyclist was not seen, or were single cyclist crashes (n=9), resulting from interactions with the infrastructure (e.g., hitting the curb, sliding on the sharp turn).

By the way, the citation links above seem to work, if you’re interested in finding out more.

What to take from this study? As with all science, more such studies are always welcome and almost always needed, but the findings here once again raise the possibility of a divergence, once again, between what looks safer and what actually is safer. Perception and psychology plays such an interesting role throughout human life, including our choices of which route to take while on our bikes.

Graphic/photo from the 2013 study