Bicycle Austin forum

MichaelBluejay

Webmaster

From: Austin, TX

Registered: 2008-05-26

Posts: 1,466

Website

Relative safety of cycling vs. driving?

How dangerous is bicycling compared to driving?  My conclusion is that cycling is 3 to 10 times more dangerous than traveling by car, on a mile-per-mile basis, as long as you don't ride against traffic.  Below is how I came up with these figures.

Some people prefer to compare risk per hour or per trip, but I prefer per mile because it seems the most apples-to-apples way:  If someone starts biking to work instead of driving, they're replacing their driving miles with biking miles.  Granted, bike routes are sometimes longer than car routes, but sometimes they're shorter too, and in any event I don't think the difference is so large as to be exceptionally meaningful.  (e.g., A bike route is rarely 40% longer than the car route.)  And yes, drivers do tend to travel farther in general than cyclists, but we can easily account for that once we know the relative per-mile risk.  (e.g., If biking is 3x more dangerous per mile than driving, but drivers travel 4x as far as cyclists, then driving is more dangerous than biking.)  In truth, all methods of comparing risk have some downside or flaw, but I think miles-to-miles is the most relevant.

Having decided to use miles traveled as the yardstick, the biggest problem we face is that we don't have good data for the number of bicycle miles traveled, because the figures vary wildly depending on which source you pick:

150 billion miles -- Consumer Product Safety Commission

6 to 21 billion miles -- Department of Transportation

6.2 billion miles -- Bureau of Transportation Statistics

I believe we can ignore CPSC's figure because it's just not believable.  (If it were true, the number of car-miles traveled would be only 20x higher.)  But what to make of the 6 to 21 billion range, which is quite a range?  Not knowing which is more accurate, I'll just use both.

784 cyclists died in 2005. That would make the death rate 0.37 to 1.26 deaths per 10 million miles.

33,041 motorists/passengers died from 3 trillion miles traveled, making their death rate 0.11 per 10 million miles traveled.

So cyclists are either 3.4x or 11.5x as likely to die as motorists, per passenger mile. Neither conclusion is very happy.

However, all these figures include people who ride dangerously, such as against traffic, at night without lights, on sidewalks, or through red lights/stop signs without adequately checking cross-traffic. A study in Washington State found that 11% cycling fatalities involved wrong-way riding. Subtracting out 11% of the nationwide deaths, we find that cyclists who don't ride against traffic are 3x to 10.2x more likely to get killed than motorists.

More than 80% of fatalities for child bicyclists 14 and under were caused by unsafe riding (riding the wrong way, running signs, etc.), However, I can't subtract out those fatalities, because then I'd be counting some wrong-way fatalities twice, since I already subtracted out wrong-way riding for all cyclists above.

I would like to subtract out fatalities where the cyclist was riding at night without lights, or riding on the sidewalk, or ran a traffic signal, but I can't find the data. If you can find the percent of biking fatalities (not crashes) caused by these things, please share!

My sources are listed in my Transportation Almanac.