There are lies, damn lies and statistics.

-Mark Twain

I suspect many of those bicycle miles are ridden in places like china where you are likely to have a large number of bikes riding in a sea of bikes.

This situation will result in quite a few scrapes and broken bones, but, it is extremely unlikely to result in death.

Compare this to a typical US commute where many of us have to do battle with the big fast metal things that like to kill us.

What I am getting at is that throwing around stats with something that varies extremely with the environment is kinda dumb.

Urban bike commuting is certainly more risky than car commuting in a similar area, because, if you are in a car in a congested area, by definition, you are likely going slow. But even at those speeds, a bike comuter can get dead with frightful ease.

Compare this to a 17 year old high school kid driving a car on rural roads to school or riding his bike to school. 17 year old boys + high speed vehicles sometimes = death. In this case, the bike is safer, IMO.

It all comes down to just weighing each situation, trying to use good judgement and crossing our fingers. If we all started living to try to maximize our statistical odds of not getting dead, we'd live pretty miserable lives.