I keep hearing on the MSN forum that painting rainbow crosswalks is a safety issue, but I can't find any data to support that.
According to U.S. Dept. of Transportation, there isn't any. That is to say, there is no evidence that one form of crosswalk marking is any more or less effective than another in achieving safety for pedestrians. What seems to matter is that the crosswalk marking color is visibly different than the surrounding roadway surface.
In my city, crosswalks are designated most often by two parallel white lines with the interior painted brick red to mimic cobblestones. Here's one of our busiest downtown intersections.
You can see the nominal red-painted, white-lined crosswalks near the bottom. Crosshatching is reserved for mid-block crossings (near the top). And at the top left is an un-reddened crossing made simply by white lines. Outside the downtown area, the simple white lines are the most common—no interior color. By the evidence, a rainbow-painted interior would actually be
more visible than plain white lines.
The safety argument seems mostly a red herring. With my engineer hat on, I can appreciate the argument that roadway markings should be clear and uniform. But community and neighborhood standards seem to be preferred, and the data do not show that this impairs safety.
Are people being run over because drivers can't see the crosswalks?
In my city, people are being run over simply because drivers don't care about pedestrians. A few years ago, Salt Lake police made a concerted effort to ticket drivers for not yielding to pedestrians in crosswalks, but the problem was never the visibility of the crosswalk. Salt Lake drivers are just jerks.
More seriously, our streets are often covered with snow. Because of that, road signs identify designated pedestrian crossings.