Automobile Traffic and Juvenile Homo Economicus

As the father of a five-year-old, this is a subject in which I have some natural interest:

Little. Yellow. Dangerous:
"Children at Play" signs imperil our kids.

Despite the continued preponderance of "Children at Play" on streets across the land, it is no secret in the world of traffic engineering that "Children at Play" signs—termed, with subtle condescension, "advisory signs"—have been proven neither to change driver behavior nor to do anything to improve the safety of children in a traffic setting. The National Cooperative Highway Research Program, in its "Synthesis of Highway Practice No. 139," sternly advises that "non-uniform signs such as "CAUTION—CHILDREN AT PLAY," "SLOW—CHILDREN," or similar legends should not be permitted on any roadway at any time." Moreover, it warns that "the removal of any nonstandard signs should carry a high priority."

The assumption here is that drivers are ignoring these signs. But, the article goes on to say, the mechanism by which the signs fail to work is not well studied:

In fact, as much as engineers know about how "Children at Play" signs do not work, one might argue there is much they do not know about how they work. As one report notes, "though these all discourage the use of such signs, none of them cites specific research demonstrating that these signs are ineffective." In other words, even if the "Children at Play" signs are a placebo, that doesn't mean they've been tested with the same rigor as a new drug.

Here's my bet: it's the Peltzman Effect. Humans have a natural risk tolerance, and will take steps to increase or decrease their risk in response to it. In other words, not the drivers at all, but the kids. Little buggers feel like the signs ought to make them safe, so they compensate with their behavior to make sure they aren't. Just like when I was a kid I kept finding higher trees to climb and jump out of until I finally found out how far I could fall without spraining an ankle.

It's also not uncommon to see "Children at Play" signs in the presence of 35 mph speed-limit signs, which is roughly akin to trying to put out fire with gasoline. It's not simply that fatality risks begin to soar at impact speeds of more than 20 mph, but that, as a study by John Wann and colleagues at Royal Holloway University in London has suggested, children, until well into their teens, are unable to detect during a normal crossing of the street the approaching speed and distance of cars above a threshold—also 20 mph. This study adds legitimacy to the increasingly popular idea, as introduced in the U.K. in 1991, that residential areas be designated as 20 mph zones.

Except that if my hypothesis is correct the kids will start wearing camouflage and dancing in the middle of the street in the event the speed limit is lowered to 20. But hey, worth a shot.