Roundabout Revolution Sweeping US
chrb writes "BBC News reports that U.S. cities are installing more roundabouts than ever before. The first British-style roundabout appeared in the U.S. in 1990, and now some cities — such as Carmel in Indiana, are rapidly replacing intersections with roundabouts. Supporters claim that roundabouts result in increased traffic flow, reductions in both the severity and incidence of accidents, and fuel savings. Critics say that roundabouts are more difficult to navigate for unfamiliar American drivers, lead to higher taxes and accidents, and require everyday acts of spontaneous co-operation and yielding to others — acts that are 'un-American.'" As a driver who's hit all of the continental U.S. states except North Dakota, I dread roundabouts and rotaries for all the near accidents (and at least one actual accident) I've seen them inspire, and have been unhappy to see them spread. Spontaneous driver cooperation doesn't necessarily need the round shape, either.
You're either trolling or ignorant... Many places I need to ride, there either are no sidewalks or they're incomplete and only exist in segments or at intersections. You'll also find that in some areas of the US it is illegal to ride a bicycle on the sidewalk. I have personally been ticketed in the past for this in my home town.
You can't take the sky from me.
However, as other posters have pointed out, there are several kinds of intersection, each with their own factors that must be accounted for, (traffic volume, etc.) and it is unclear if the studies have taken into account the differing characteristics of the intersections that were replaced.
Additionally, the IIHS, for one, considers roundabouts distinct from rotaries. Since many others do not make this distinction, it is difficult to tell what kind of traffic circles were studied, and what kind of traffic circles are being installed in the US's "roundabout revolution."
Either way, the knee-jerk reaction of "rotaries are dangerous" at the least needs a conditional and at best is quite false.