Advocacy & Safety - Scientific American -- The more the merrier!

http://www.sciam.com/podcast/episode.cfm?id=46B9E06D-FC8E-9007-E864EF0B7227869B&SID=mail&sc=emailfriend

I found this to be very true in comparing safety in Seattle and Los Angeles. The bike infrastructure gets folks out on their bikes, and all of the cyclists, regardless of infrastructure, make it a safer cycling city.

I don't doubt Jacobsen's numbers. (The Safety in Numbers theory for bicyclists comes from an article by Peter Jacobsen.) But I do have doubts about the massive assumption that is made to explain them -- that the correlation between greater numbers of cycle-commuters and lower collision rates (Jacobsen used census numbers for commuters and compared them to total car-bike collisions in different communities) is caused entirely by a change in motorist behavior. Let's think about this for a second. Do we really believe that a gain in ridership from, say, 1% to 2% of the commuting population would cause such a massive improvement in motorists' noticing and avoiding cyclists? I don't think so. What Jacobsen overlooks here is that the community with the larger portion of cycle-commuters is a community with a relatively large number of safe, conservative riders compared to the cycling population at large in any city, which will include all the familiar characters, the drunks, the kids riding out of their driveways to disasters which are then included in the collision count. When you have lots of commuters you change the overall makeup of the cycling population. Commuters come out and ride more miles, better. The percentage of child-bicyclists among all the city's bicyclists drops precipitously. The cycling population overall does indeed become safer, more prudent and aware on the streets, and this would probably be reflected in the accident rate. But the drivers don't magically figure out how to drive just because they're seeing 10 bicyclists instead of 5.

IOW, your safety is still up to the dude in the mirror. Not the dude behind the windshield.

I don't doubt Jacobsen's numbers. (The Safety in Numbers theory for bicyclists comes from an article by Peter Jacobsen.) But I do have doubts about the massive assumption that is made to explain them -- that the correlation between greater numbers of cycle-commuters and lower collision rates (Jacobsen used census numbers for commuters and compared them to total car-bike collisions in different communities) is caused entirely by a change in motorist behavior. Let's think about this for a second. Do we really believe that a gain in ridership from, say, 1% to 2% of the commuting population would cause such a massive improvement in motorists' noticing and avoiding cyclists? I don't think so. What Jacobsen overlooks here is that the community with the larger portion of cycle-commuters is a community with a relatively large number of safe, conservative riders compared to the cycling population at large in any city, which will include all the familiar characters, the drunks, the kids riding out of their driveways to disasters which are then included in the collision count. When you have lots of commuters you change the overall makeup of the cycling population. Commuters come out and ride more miles, better. The percentage of child-bicyclists among all the city's bicyclists drops precipitously. The cycling population overall does indeed become safer, more prudent and aware on the streets, and this would probably be reflected in the accident rate. But the drivers don't magically figure out how to drive just because they're seeing 10 bicyclists instead of 5.

IOW, your safety is still up to the dude in the mirror. Not the dude behind the windshield.


Good points Robert.

If people are interested in the article ... (http://www.tsc.berkeley.edu/newsletter/Spring04/JacobsenPaper.pdf)

...
If people are interested in the article ... (http://www.tsc.berkeley.edu/newsletter/Spring04/JacobsenPaper.pdf)

Jacobsen should have compared numbers of commuting cyclists with the collisions of commuting cyclists, or at least of adult cyclists. It seems obvious in retrospect, but I can't see that kids' collisions have been culled from his data.

I think people really, really want to believe this paper's conclusions about greater numbers of cyclists fundamentally changing motorist behavior/awareness. It's a simple answer that points to a simple path for advocacy (just attract more riders!). But does anybody notice that it also takes the onus (love that word) of cyclists' safety off the cyclists and puts it squarely on the drivers? The cycling population can't change its behavior, J. assumes, but drivers will, like magic, and all we have to do is increase the number of riders. Very interesting. Cyclists tend to shrink from responsibility for their own safety, and everybody loves an easy answer. But I'm afraid that's not the way it works.

Another thought: Safety in Numbers works for bicyclists and walkers, but not for motorcyclists? As motorcycling has become wildly popular over the past decade, fatality-per-registration has held steady and fatality-per-mile numbers have almost doubled.

Another thought: Safety in Numbers works for bicyclists and walkers, but not for motorcyclists? As motorcycling has become wildly popular over the past decade, fatality-per-registration has held steady and fatality-per-mile numbers have almost doubled.

Seen recently in Texas...

Another thought: Safety in Numbers works for bicyclists and walkers, but not for motorcyclists? As motorcycling has become wildly popular over the past decade, fatality-per-registration has held steady and fatality-per-mile numbers have almost doubled.

Unless my memory is failing -- with baby, don't be surprised if only mush is left -- he also doesn't control for the number of cars on the road either. But I think you got the big piece of the "pie". That is, a large portion of the effect is simply due to "different" people cycling.

Also, even though an entire road system doesn't change rapidly, where people choose to drive/cycle can be quite selective thus decreasing the interaction and number of accidents.

My own model of the world suggests that there is some type of feedback effect due to some combination of driver and cyclist "education". That is, drivers learn how to interact with cyclists. Since cyclists still drive, as drivers they become more aware/sympathetic for cyclists. But I would be surprised if the marginal effect is large. Partially since I think cycling is relatively safe (to driving) activity in the first place.

I am not up to date on the motorcycling statistics. But in the case of motorcycles, perhaps it is due to a demographic change in who motorcycles. Anecdotally, it seems like older people are hopping on the hogs nowadays. Older people might be more likely to die from a motorcycle accident. And I recall that most accidents occur with "new" motorcyclists. So an increase in popularity would result in a spike of "new" people on the road.

...
My own model of the world suggests that there is some type of feedback effect due to some combination of driver and cyclist "education". That is, drivers learn how to interact with cyclists. Since cyclists still drive, as drivers they become more aware/sympathetic for cyclists. But I would be surprised if the marginal effect is large. Partially since I think cycling is relatively safe (to driving) activity in the first place.

I am not up to date on the motorcycling statistics. But in the case of motorcycles, perhaps it is due to a demographic change in who motorcycles. Anecdotally, it seems like older people are hopping on the hogs nowadays. Older people might be more likely to die from a motorcycle accident. And I recall that most accidents occur with "new" motorcyclists. So an increase in popularity would result in a spike of "new" people on the road.

I suspect that is correct -- that there is some small effect from motorists due to more cyclists on the streets, but that the main effect driving the differences in accident rates among bicycling populations comes from the differences in the populations themselves. And the composition of the motorcycling population has changed profoundly, with more rookies than ever. Experienced bicyclists and motorcyclists have astronomically lower per-mile injury rates than beginners. But I guess that is not much of a feel-good story.

I've just spent almost 4 months away from the urban environment that I normally ride. I've been riding on rural and dirt roads with little or no real "traffic". Isolating the bike rider from an autocentric environment provides a great opportunity to see just how much we are responsible for our own safety. Dealing with steep descents, rough roads and other road hazards is good reminder of the kind of constant attention bike riding requires even without cars added to the mix.

Returning to the urban environment complicates the issue. While numbers of cars and numbers of bikes must be taken into account I think speed is a greater factor than volume. I wonder if a greater number of cyclists contributes to slower auto speeds on those roads and may be, in part, responsible for greater safety.

I do find that city streets with more cyclists have more bike-friendly motorists, and there are fewer attempted door prizes (I even see people looking out their windows and back for bikes:)), than on streets with fewer bikes.

However, it is not certain that there is a cause-effect relationship here, but it does seem that there is a positive correlation.

BTW the streets I'm thinking of (Bloor, Harbord, St. George, and Beverley in Toronto) have over 10% of road users as cyclists at rush hour (depending on the time, Harbord might be closer to 50-75% cyclists by numbers, as many bikes pass by cars that are stuck at lights, and, of course bikes are smaller and nimbler than cars).