Thursday, October 12, 2017

Speed limits

John Lower wants to redo our speed limits, based on measurements by smart sensors in traffic control systems. He thinks that might make a lot of speed limits slower.

Redoing the old studies from the 50's and 60's makes sense. Adjustable speed limits does not: keep the number of driving variables small if you want minimum confusion. For example, reversible lanes are OK if there are physical barriers that shift, but if they are controlled by overhead X or O symbols you are asking for accidents. "Motor memory" of how fast I am supposed to be going along street-type X plays a role in my driving style, and I suspect I'm not alone.

And if photographic speed limit enforcement starts to become ubiquitous, I expect a significant rise in vandalism. Air rifles and slingshots and paintball guns would probably be useful for the purpose.

Mike Royko had something to say about speed limits.

UPDATE: "Unknown" points out in the comments that the variable speeds can work.

1 comment:

Unknown said...

The M25 around London has variable speed limits (with strict enforcement now including photo enforcement not just of instantaneous speed but of average speed between cameras). The point of this is not safety, but maximum thru-put of vehicles -- if you can slow the arrival of cars to a constriction point, the backup may never begin and a traffic-jam can be avoided. It seems to work there, but the infrastructure costs of the countless overhead gantries reminding you of the current speed must be enormous!

It seems to work.

Other than US interstate highways, where the trend I've seen has been raising the speed from arbitrary energy-motivated 55 to 65 and then 70 or 75, the changes in speed-limits that I've noticed on familiar roads in the UK and NE US have mostly been down - suburban/small-town through-routes that were 45 or 50 MPH are now marked as 35 or 40.

This has not worked - my observation is that the typical speed of vehicles on these roads has not changed a bit, except in the vicinity of regular speed-traps in the US and speed-cameras in the UK.

I found it fairly easy to find more recent studies -- is one from the mid-1990s that looks at how the number and severity of accidents changes as speed-limits are increased or decreased.

As far as I can see the "85% rule is not followed as much as authorities claim. Partly that is because new roads are engineered with "design speeds" that it makes sense to post, and speed limits are altered on account of accident history and local conditions such as the mix of cycle and pedestrian use. States and localities will have policy that "roads of X type have a speed limit of X" which makes all housing-development roads a uniform 25, for example, and this arbitrary limit is rarely near-to-or-above the 85th percentile.

Since learning of the 85% rule a decade-and-a-half ago, I've become an observer of where the speed-limit changes on roads, and I try to work out why the speed-limit on one one section of road might be 45MPH but 30 meters later 35MPH is specified. Often it is because of a jurisdictional change, very obvious when a rural-road through farm-fields intersects the corner of a political division and there is a different speed limit for that short section even though every other condition of road, surroundings, and traffic is identical. In one place I never understood a 35MPH section on a stretch of otherwise 45MPH state-route, until I visited a garage-sale on a side road and found upon exiting that road that the limited visibility because of a curve and hill made safe entrance to the state road nearly impossible.

I suspect highway departments use the 85% rule as a mantra to put-off annoying neighborhood activists -- my wife is constantly complaining about vehicles speeding on our road, and my observation when she says "that car/truck was going way too fast" is that my estimation of the vehicles' speed by visual observation is at or within 5 MPH above the posted limit, but it is noisy because of lugged tires or use of a low gear.

Anyway, I stipulate that the claims of the article you link are wrong:
• Speed limits aren't too fast because of the 85% rule, as the 85% rule is only one of many criteria used for setting the limit. If local conditions (such as presence of cycle commuters such as himself) make a lower limit appropriate, authorities have the power to make the limit lower.
• The Solomon studies behind the 85% rule are not obsolete from being old. There certainly have been studies showing that higher speed leads to more and more severe accidents, but that is a distinct problem from the appropriate setting of speed limits. And there is plenty of recent research to be found that replicates the Solomon studies. It is an error to think that lowering the speed-limit (in isolation) will lower the number or severity of accidents on a road.