This factoid jumped out at me from the Seattle Department of Transportation’s 2010 Traffic Report. Guess what the weather was like when 71 percent of all car-bicycle collisions took place? Clear or partly cloudy. Fourteen percent of collisions occurred when it was overcast. Only twelve percent when it was rainy.
You just assume that the weather is bad, it’s slippery, there’ll be more accidents. But no. In Seattle, looking at the absolute numbers, high season for bicycle-related accidents for the past five years are the months May through September, with April and October as the shoulders. That makes sense in one way because there are simply more cyclists out and about when the weather is nice.
But consider the statistics for pedestrians, who are also, you assume, out and about when the weather is nice. Seattle is flush with tourists not looking where they are going each summer. Yet the pedestrian high (-chance-of-being-hit) season is the months November through January. Again, common sense, except you might apply the same common sense to bicyclists (weather’s bad, it’s dark, can’t be seen) and be wrong.
Nor does clothing visibility seem, in this batch of statistics, to offer much of an advantage. Of the collisions where the bicyclist’s clothing was noted, 35 were wearing light or reflective clothing compared to 42 wearing dark clothing, and 122 who were wearing “mixed” light and dark clothing. The lesson seems to be to go bright or go ninja, but don’t hedge your bets.
Most dangerous day of the week? Wednesday. Least dangerous? Sunday (I would guess simply because of lower traffic volumes). Most dangerous hours of the day? 8 to 9 a.m., and 3 to 7 p.m. The leading age group for accidents is 25-34.
The leading reason a driver hit a bicyclist (142 times) was given as failure to grant the right of way. But before cyclists get their chamois-padded bike briefs in a twist, consider this: in 66 collisions, the bicyclist failed to grant the right of way to a pedestrian. The collision was most likely to happen at an intersection (60 percent), and another surprise, more likely to happen when bicyclists were riding with traffic (32 percent) than entering or crossing traffic (18 percent).
That said, SDOT’s data can be surprisingly incomplete. In the last instance, 45 percent of the time, no one knew or wrote down what the cyclist was doing–a startling omission given that these are car vs. bike collisions. The age of the bicylist was undetermined 21 percent of the time. 45 percent of the time the “facility type” (e.g., roadway, bike route) was missing.
I would also take issue with SDOT’s assertion that “the citywide count showed a decline in bicycling” of 15 percent. A one-day count is really only useful for establishing the presence of something. Whatever else is true, citywide, some 3,961 people biked around Seattle on a particular day in 2010. But you can’t be sure, by comparison solely to previous one-day counts, whether you’re really seeing an increase or decrease. That’s true as well of the 20 percent “uptick” in bike commuters to downtown, of 3,251. Maybe it is an increase. But really, it’s more important to know that 3,251 people biked to downtown. So when people tell you how impossible it is to commute to downtown on a bike, you have 3,251 comebacks.
The wider the road, the wider the bike lane, is what makes it safe. Visibility, the earlier in the morning is better, no-no on Friday and Saturday nights, and of course defensive riding has worked for me, knock on wood.
Sorry, but there is zero information in these ‘stats’ regarding the wisdom of wearing different colored clothes. The key question is the proportion of dark-wearers vs light-wearers vs mixed-wearers.
Let’s say 42 bicyclists in Seattle wore dark, and 35 wore light (and everyone else wore mixed). In other words, 100% of the solid-color-wearers would have been hit, and a tiny percentage of the mixed wearers. The point is that without knowing the mix of clothing, the absolute numbers are useless.
But of course, as Dan says – defensive riding is critical. But don’t get complacent!
You are correct, sir! But that doesn’t get you a ninja joke, does it? So there are tradeoffs. Yes, to really tell if clothing helps much, you need to know the percentage wearing light/reflective gear in the total biking population. It would also help if you could screen out instances in which cyclists were at fault, in which case clothing is likely to be irrelevant.
On top of that as regards clothing, I read a safety tip somewhere that if your route has you coming out of the rising or setting sun (i.e., the sun is at your back) you are most visible if you’re wearing bright red or bright orange, but definitely not white. (I’m assuming that blends into the sun’s direct rays so you disappear.)
So add in your cardinal direction and time of day to the clothing visibility calculation if you want to mess with the real meaning of the stats some more.
Thanks for the post (which I found thanks to Cascade Bicycle putting it out via Twitter). I’m now going to go datamining for Spokane, where I ride in the high-vis combo of regular business attire–usually a skirt–and heels. Seems to work as well as lemon yellow or lime green, at least so far.
barb
Nice Post, Michael!
One thing that I’d want to correct is the conclusion that more people on the road means more crashes. This is not necessarily true– in fact, we’ve seen from some cities that the exact opposite is true. Take NYC, for instance: http://transalt.org/files/campaigns/bike/images/ridership_graph.jpg. More riders on the road make it safer for everyone– including drivers. And more bike infrastructure also translates to safer streets for bikes– and cars and pedestrians.
Portland tracks similarly to NYC. Portland actually saw nearly a decade’s worth of zero bike fatalities after dramatic increases in ridership. And it’s probably the reason why Seattle’s Bicycle Master Plan calls for a tripling of bicycling with an overall crash reduction of 1/3.
As for the overall ridership numbers in Seattle, well, you’re spot on. There’s been quite a bit of press about that and I feel like we’ve all helped explain it pretty well. By most other measures, bicycling is rising steadily and we’re up to 2nd in the nation for commute mode split. What’s clear, too, is that the need for more and better data collection is in order.
Thanks again for your analysis and for posting–
best,
John at Cascade