For your consideration, a Freakonomics post linking to an interesting bike-car crash study compiled via Toronto police reports.
Short version: ~90% are caused by motorist action.
That doesn't make me feel any better. I'd rather they found that half of all accidents are collisions with stacks of goose down pillows piled 10 feet deep on the highway.
The corresponding stats on motorcycles are a little better. In that, 3/4 of all accidents are from passenger vehicle error.
3 beers + bicycle ride is much safer than 3 beers + motorcycle ride
There are apparently some methodological problems with that report. I don't know if they are mentioned in the Freakonomics comments. I've been disappointed with the Freaks lately- some of the posts have contained serious, obvious oversights. They seem to be chosen for freak value rather than academic rigor 
Long story short - it's not hard to believe that 80% of all collisions involve error on the part of BOTH parties.
Longer story - if a cyclist is riding on a sidewalk against traffic at high speed and shoots into the crosswalk and gets hit, it's the motorist's "fault" but it's still stupid for the cyclist to do that. If a cyclist is riding in the door zone and gets hit, it's the motorist's fault - legally and morally - but it's still avoidable and counter to all good advice. If a cyclist is hugging the curb and a motorist tries to squeeze by, and either misjudges, or the cyclist wobbles, it's absolutely the motorist's "fault", but once again, avoidable.
I guess what I'm saying is that reports like this do cyclists a disservice if they create the impression that cyclists have no reason to operate responsibly, and thus encourage more crazy wrong-way sidewalk riding.
I won't even have 1 beer and get on my motorcycle. I will have beers and ride my bicycle.
I think there are plenty of flaws with this report, but my observations lead me to believe the percentages are probably pretty accurate.
if you're at the point where you even care about this report, then you're probably operating a bike in a pretty responsible way to begin with
Yah, but I doubt that applies to most of the people reading the Freakonomics blog.
