index previous comment next comment

Roundabouts have been shown to reduce intersection injury crashes by 76% compared to signals, and to reduce fatalities by 90% (Retting et al, 2001). Results of these American studies, published by the Insurance Institute for Highway Safety, and by the American Journal of Public Health, are identical to findings from decades of overseas research. Put another way, comparable crossroad intersections have FOUR TIMES AS MANY INJURIES - including severe brain and spinal cord injuries - and about TEN TIMES AS MANY FATALITIES as a roundabout. Where roundabouts have replaced signals, repeated studies have shown they reduce pedestrian accidents by 30-50% (Lalani 1975, Daley 1981, O'Brien 1985).

Unfortunately, these safety facts were not emphasized in Dr. Richard Long's report to the Access Board: a report which was used to develop the Access Board's current design proposals. The proposed new unfunded federal mandate would require complex signals and barrier systems at every roundabout crosswalk, regardless of how easily the crosswalk operates or whether a blind person ever uses it. At $100,000 or more per intersection, that's an expensive proposal, and since it has serious ramifications on public safety, it bears close examination.

Like more than 90% of US crosswalks, most roundabout crosswalks are not signalized. Because a pedestrian refuge is provided mid-crossing (shortening the crossing distance), and because vehicles operate at unusually low speeds (typically 15-20 MPH), the overwhelming majority of roundabout crosswalks are extremely easy to use, and like the vast majority of all crosswalks in the US, they simply don't need signals. In such locations, even if signals were provided, pedestrians wouldn't use them.

Traffic engineers have known for decades that, if unwarranted and infrequently used, signals can confuse drivers, become ineffective, and INCREASE accidents, causing more injuries to pedestrians and vehicle occupants alike. That's bad for public safety, but too few lay people understand it. People tend to think that signals always improve safety, but signals can increase speeding, distract drivers' eyes away from traffic and pedestrians, and create a false sense of security for pedestrians. Signals do not put a concrete wall between vehicles and pedestrians: pedestrians are struck at traffic signals with sickening regularity.

In locations where pedestrian and traffic volumes warrant them, crosswalks should have signals, and many roundabouts in the United Kingdom and Europe have pedestrian signals. They're common in London, Birmingham, and other cities. Examples of appropriate locations for signalized crosswalks include high-volume urban roundabout crosswalks, and locations where pedestrians (both blind and sighted) are most frequent. These factors are easily quantifiable. At rural intersections or low-volume locations, and locations where pedestrians are infrequent, signals are not used because they confuse drivers and unnecessarily increase highway construction, maintenance, and operating costs.

To assure that traffic signals are only installed where prudent, "warrants" have been developed for traffic signal installation in the United States. In the United Kingdom - a country with decades of experience with roundabouts in a wide variety of locations - the warrant for a signalized pedestrian crosswalk at a roundabout is where PV squared is greater than 1*10^8 (in words, where the number of pedestrians, times the number of vehicles per hour squared, exceeds a value of 100 million). Use of appropriate signal warrants assures that signals are provided only where needed, and not where they are unnecessary and potentially harmful.

The ramifications of an ill-considered intersection design policy can negatively impact the general public in unintended ways. For example, if ALL roundabout crosswalks were required to have signals, about $100,000 would be added to the cost of each roundabout, making them unnecessarily expensive in comparison to other intersection alternatives. As a result, far fewer roundabouts would be built, and many more of the common alternative - a signalized crossroad - would be built instead. As stated previously, studies show these have FOUR TIMES as many injuries, and TEN TIMES as many fatalities as a roundabout. The United States currently has about 15,000 deaths and about 1 million injuries at intersections every year, and installation of well-designed roundabouts might prevent countless human tragedies. Meanwhile, signals at unwarranted locations may help no one at all, and could in fact be harmful.

A single-user approach to traffic engineering would be a mistake. All users of an intersection must be taken into account, and the appropriate solution needs to be provided that will provide the greatest benefit to the public in each specific situation. No one wants more people injured or killed because we impose an ill-considered intersection design policy.

The Access Board proposal to require signals at all roundabout crosswalks needs to be reconsidered. Signals should be installed where they are warranted, and where there is no better alternative. In specific locations where users have special needs, the needs should be evaluated and provisions should be made in the design.