Speed limits don't work

I was raised to be a very concientious driver. My father, who did the bulk of the driving on any road trip, follows every law to the letter, never exceeds the speed limit (actually, he did once - it was late, and dark; he was a bit agitated because we had taken a wrong turn somewhere; and he exceeded the posted limit by three or four miles per hour - I could see the speedometer form my place in the back seat) He also makes full and complete stops at every stop sign. And as a starting driver I emulated this behavior. But, as is the case with most drivers, I started to relax a bit with time. I'm a little older now than when I was in college, and I suppose that's part of the reason why I don't drive as fast as I used to. The cost of gasoline is another - mileage plummets with excessive speed. But the road itself also plays a major role.

I realized this several months back while driving to the store. The road that I take most of the way starts out in a residential community, with a 30 mph limit, then opens up into a thinly-developed area. Here, the north- and south-bound lanes are separated by a grassy median and the lanes are wider; here the speed limit is 55 mph. For the most part, the road is straight in both areas. But the thing is this: people drive at about 40 or 45 for the whole distance. For the 30 mph zone this isn't a surprise - people seem to drive between 5 and 15 mph over the speed limit around here. But driving 40 in a 55 is not common around here in the absence of very heavy weather.

I shared this observation with someone - I now forget who, I'm afraid, and was asked in return "does it feel like a 55mph road?" It doesn't, of course. There may be a grassy median, and the lanes may be a bit wider than typical for residential areas, but it lacks the guard rails and other attributes of a parkway, far less an Interstate (did you know that Hawaii has an Interstate highways? Think about that. Interstate doesn't mean interstate.) Not long after, I found a (PDF) report of a study on, among other things, the relationship between posted speed limits and the actual speed of vehicles (see the section entitled "Operating Speed and Posted Speed Relationships", starting on page 46). Often, the 85th percentile (that is, the speed at or below which 85% of vehicles are travelling) was greater than the posted speed limit. This comes as no surprise. What was enlightening was that the relationship between the posted speed limit and the 85th percentile seemed to vary, in part, based on the type and setting of road: rural vs suburban& urban (S/U).

Rural situations tended to see a greater percentage of drivers traveling at or below the posted limit. This was especially so for two lane rural highways. S/U roads showed a lower percentage of vehicles at or below the (posted) limit, and showed a greater variation in the relationship between the 85th percentile and the posted limit. Something about the nature of the road influenced drivers in their choice of speed (yes, yes, there are plenty of variables that I'm ignoring - read the paper yourself and then come back and argue).

In a later section of the same paper ("Changes in Posted Speed Limit", p53), the effect of changing a speed limit is discussed. Of the sites evaluated, typically the initial speed limit was at the 20th percentile, while the new limit was raised to the 43rd percentile (of the original data). Following the change, the rate of driver compliance rose, suggesting that the change is speed limit influenced drivers - but that's not the case: average speeds, the standard deviation of the speed, and the 85th percentile generally changed by less then 2 mph each. Compliance rose not because the drivers followed the posted limit, but because the limit followed the drivers.

So, it would seem that there is something besides the speed limit that influences the speed of th typical vehicle, which brings me back to my drive to the market: the driver's experience of the road in the 30 mph section is very similar to the driver's experience in the 55 mph section. The grassy median is not a highway/parkway-exclusive feature; it is seen on boulevards as well. Further, it is not a standard highway feature - many highways lack a median, grassy or not (particularly in the city).

I also found the shared space philosophy (look it up on Wikipedia, and follow the references). It appears that removing roadway signs, traffic lights, cross walks, etc results, in some cases, in a reduction in the number of accidents. Personally, I think it has to do with drivers surrendering their responsibilities to the traffic control devices around them - removing the devices forces the drivers to assume personal responsibility. But whatever the case, this suggests that whatever effect speed limit signs have, it isn't a good one.

[boldface formatted 4 Oct 2007]