Bicycle Crashes in DC

The District of Columbia keeps all kinds of statistics on crashes - who was involved, where they happened, etc...but the statistics have numerous well-known flaws, especially where it pertains to cyclists.

Not all traffic accidents are reported, and even fewer involving bicycles are. Most accidents reported involve an injury of some sort, but the smaller accidents often don't get into the system.

The forms police officers use are not designed to gather important information about bicycle accidents - like was the rider wearing a helmet. In DC, officers still use a single paper form for all accidents and so to add even one line increases the complexity. In some other cities, officers fill out a digital form. Responses to earlier questions determine the later questions to be asked. (So if the police officer answered "Yes" that one party was a cyclist, then it will ask all of the bicycle questions. Otherwise it will skip them)

Also police officers often fill in the closest intersection to determine where the accident occurred, even if it was on a bike trail. Or if the accident was midblock (like where a parking garage exit is). So that throws all of the geodata out of whack.

Still, there are some good data in the reports they create. Here's one map from the report using 2000-2002 data. Here are the intersections with 5 or more crashes from 2000-2002:

Looking at the map 14th Street seems to be the location of more crashes than any other street.

Comments

[Keroicap]

Few thoughts:

1. It's important to keep in mind that these crash numbers aren't normalized for the number of cyclists travelling through that location. The report indicates that the most collisions took place at Connecticut & L and 13th & I, and then calls these "the most dangerous intersections for cyclists".

Well, maybe. Or maybe there were just more cyclists using those intersections. To use a ridiculous example to prove the point, the Beltway presumably has zero bicycle-auto collisions in any given year. Does that mean the Beltway is a safe place to cycle?

2. These crash data would be a lot more helpful if they indicated whether the motorist and/or cyclist were cited by the police. Then we could actually guess at how many crashes are the fault of the cyclist and how many the fault of the motorist. It's pretty hard for a bike and car to collide without one or the other breaking the law, I think.

3. Interesting that 72% of the auto drivers involved in bike collisions were male. Anyone know if that's representative of auto-auto collisions as well? Seems hard to believe.

[Keroicap]

My previous post was referring to the older report (97-99). The newer one (00-02) seems better -- nothing about "dangerous" interesections, and there seems to be more consideration of "fault". Not that I necessarily trust a DC police officer to correctly assign fault in an auto-bike collision.

Seems like there are a lot of left-hook and right-hook crashes, since a large majority of cyclists are going straight at the time of the crash, and only a plurality of motorists are going straight.

Finally, someone needs to create a form that doesn't refer to bicyclists as "pedestrians" unless they're walking. It's both confusing, as you can see in the report, and legally inaccurate given that we qualify as vehicle operators.

[contrarian]

"Interesting that 72% of the auto drivers involved in bike collisions were male. Anyone know if that's representative of auto-auto collisions as well? Seems hard to believe."

This site:

http://www.cars-suck.org/littera-scripta/LAB-talk.html

quotes a study of road fatalities of bicyclists in New York City over a four year period (which I have not found published). Where the identity of the driver was established, 97% of the drivers were male.

"The grossly disproportionate number of male cyclist-killers strongly suggests that driver aggression (and not just cyclist impulsiveness or incompetence) plays a significant role in killing bicycle-riders in New York City — a finding confirmed in a later report showing that the foremost behavior that’s killing cyclists is aggressive passing by drivers."

[contrarian]

Here's the link to the study, "Killed by Automobile."

www.cars-suck.org/research/kba_text.pdf