I think something missing from this study (and previous ones, such as the San Diego study that suggest a “false sense of security” in marked crosswalks) is consideration of why motorists fail to yield to pedestrians in crosswalks and whether a policy of generally not marking crosswalks, as followed in San Diego and many other communities for 30-some years after the oft mis-interpreted study there, breeds a false sense of entitlement on the part of motorists (i.e., lack of marked crosswalks reinforces the motorists’ belief that pedestrians don’t belong in the road and their ignorance of the presence of unmarked crosswalks; it is really rather fantastic to expect that motorists, who ignore such posted direction as speed limits and stop signs, would somehow pay attention to invisible demarcations for pedestrian safety). More specifically, I’d like to see a study of the potential “halo effects” of comprehensive marking of crosswalks within a city. Such an effect has been observed with photo enforcement of traffic signals; it is reasonable to expect, and anecdotal evidence in “pedestrian friendly” cities suggests that the more obvious the pedestrian infrastructure, the better yielding behavior by motorists.

The other, rather obvious consideration, is that the faster traffic is moving, the less likely motorists are to yield to pedestrians, as it’s just more difficult and disruptive to stop from a high speed than a low one. The indictment should not be of marked crosswalks, but rather of high-speed roads in pedestrian environments.

As the Zegeer study points out, the behavioral aspects have not been well studied. Rather, previous researchers have focused on speculation about the behavior of pedestrians (the unsupported “false sense of security” of the victims), while almost completely ignoring the behavior of the motorists (the ones doing the killing).

http://www.tsc.berkeley.edu/newsletter/summer2006/zegeerxwalks.html