In the comments of yesterday's post on DDOT released bike data, John Cruz expressed surprise that crashes haven't dropped as bike lane miles went up. I pointed out that if crashes remained the same, while ridership went up, that meant the streets were getting safer.
Using only DDOT's data I made the following graph.
Some caveats:
- It uses commuter data as a proxy for trips, which may not be perfect.
- It uses percentages instead of actual counts. I could get the actual counts if I just knew how many commuters there were in each of those years - which is probably in census data, but I don't have time to look it up
- I combined 2001 bike lane and crash data with 2000 bike commuting numbers, to cover a longer period.
Still I think it shows that DC is becoming safer for cyclists. It does NOT show that bike lanes are the cause of this. There could be any number of explanations. But bike lanes could be a reason or even the main one.
GIGO. I can't think the data is that good.
Generally speaking, I'd agree with the conclusions. Cycling in DC is pretty safe. Which is why I'm mystified by the need to criminalize more accidents, throw more liability insurance on motorists and enforce helmet laws.
Posted by: charlie | August 25, 2011 at 11:33 AM
Which is why I'm mystified by the need to criminalize more accidents, throw more liability insurance on motorists
Because many accidents are caused by a criminal level of negligence and people in general don't have enough insurance to cover the costs of killing someone. If killing someone is rare, than the cost of the liability insurance for it should not be large.
Which part of the data do you think is not good. Miles of bike lanes seems pretty accurate. Crashes reported is pretty easy to determine and bike commuting is done by survey. Since the actual number of crashes or bike commuters is irrelevant (only the ratio matters) the trend should be pretty realistic.
Posted by: washcycle | August 25, 2011 at 11:38 AM
Bike commuting, as we've talked about before, is a bad proxy for bike riding. Those surveys are way off.
And I doubt the crash numbers are at all accurate either. They perhaps account for 25% of all "bike-crashes", mainly the very serious ones. Plenty of others out there.
GIGO.
Posted by: charlie | August 25, 2011 at 12:01 PM
True, but all we want to know is the change. I think it's pretty reasonable for the ratio of bike trips to bike commuting to remain relatively constant, as well as the ratio of unreported to reported crashes. So we'd end up with larger numbers, but the same basic trend.
Posted by: washcycle | August 25, 2011 at 12:09 PM
But the change in rates in only relevant if components of the ratio still have the same meaning. Here the type of rider and riding has changed. For instance, if new riders simply ride slower I'd expect less reported crashes. But this has little to do with the environment being safer.
Posted by: Geof Gee | August 26, 2011 at 12:52 PM
the type of rider and riding has changed
And why is that?
Posted by: washcycle | August 26, 2011 at 01:08 PM