How healthy is bike commuting? More than you might think.

An article published in the Washington Post today alerted people - on the cusp of bike to work day - that bike commuting might be more dangerous than they think (how do they know how dangerous I think bike commuting is?). The article, it will shock you to hear, has some problems but first and foremost is the general thrust.

Let's just take all the primary elements to be true. That cyclists die at twice the rate of motorists. That cyclists suffer much higher exposure to air pollution. That despite all this, the physical and mental health benefits mean that regular bike commuters live longer than those who commute by car.  Which of these facts do you think is the most important one? For me it is easily the latter, since it is stated with consideration of the other two, which feed into it. It's the final conclusion of all of it, and it would be the bottom line that most people care about. What are the health implications of my choice? I'm sure that choosing to exercise, in general, increases your chances of accidental death, but the Post would never run a headline than read  "running is less safe than sitting on your couch." [Ok, these days, the Post might] Which is why I would have written the headline above instead.

What Sadie Dingfelder did was the equivalent of the article below

Maryland Schoolteacher his with $50 million tax bill

A Maryland schoolteacher earning $42,000 a day was suddenly hit with a $50 million tax bill yesterday after winning the $340 million Powerball jackpot....

That's all true, but the presentation is somewhat deceptive. The least important fact is left until the end, which is fine for an M. Night Shyamalan movie, but not journalism.

Data doesn't  compare the same types of people

But we shouldn't take the facts stated above to be perfectly true - at least not as interpreted and applied to the question on the relative safety of bike commuting. The Beck study compares the average person biking in the average place to the average motorist, and as such doesn't tease out bike commuting in the city. It doesn't answer the question "What are the safety impacts of switching from driving to biking for my commute."

Comparing the average type or person is flawed, in part, because the population of motorists and cyclists differ so much. Cyclists are overwhelmingly more male and men are fatally injured 122% more often while cycling than motoring. Women on the other hand are only 15% more likely to suffer a fatal injury. This might be because men are more likely to be involved in recreational cycling, riding a road bike on a county highway for long hours when fatigue begins to play a role. And recreational cycling isn't really part of the whole "commuting vs. driving" question.

In addition, there are many cyclists under 16 years old, but very few drivers of those ages. And children are about 4 times more likely to be killed while biking than while driving. And that relates to another issue, the fatality rate for motoring is pulled down by the fact that some motorists are children riding in car seats or booster seats, which makes them significantly safer. But you aren't going to commute to work in a car seat. And many passengers are in the back seats, where you are unlikely to commute, which are safer as well. The data doesn't tease out how safe driving is or even riding in the front passenger seat - which matters if comparing commuting methods.

And the two complications of age and gender mask one another, If we look at the data broken down by gender, the male cyclists might be much younger than the male drivers - in which case more fatalities might be expected because young people take more risks. If we compare by age, then we get a group of cyclists that consists of a higher percentage of men than the drivers do, which again is a more risk taking group. So in each case we're comparing a more risk taking group to a more risk adverse group.

Finally, these groups aren't randomly selected. They're self-selected. I'd expect the group that bikes to be less risk adverse than the group that drives, and thus a higher fatal injury rate among them.

In other words, the data presented here doesn't tell you if YOU personally will be more at risk if you bike than drive. It tells you that the self-selected group of people who choose to bike are more at risk than a self-selected group of motorists. 

As an example of how this skews things, if you divide the total annual injuries reported by NHTSA by the commute rate by car and bike as reported by Census, cyclists are only injured 50% more often than drivers. Meaning if we look more closely at just commuters, the relative risk goes down considerably. 

Exposure is tricky

The exposure is always tricky. The denominator used in this study is trips, but time spent travelling matters too. The tree across from my house has gone on zero trips for zero miles, but it's been involved in two car crashes (that I know of). Exposure time matters. Day long leisure bike trips can throw the whole thing out of whack. 

 

Nor does it account for time of day differences. There is way more bike commuting between 2am and 6am than there is transit commuting, for example. 

 

Where matters too

And, as the article points out, where you bike matters a lot. Cities are likely safer for cyclists than county roads due to lower speeds. But on the flip side, highways are safer than other roads for drivers. So if you live in the city, the risk of cycling will be much different than if you life in a rural area.  And even then, as the article points out, not every city is the same. This difference in the types of facilities on which people drive and on which they bike is another limitation in the data. It means it doesn't tell me if I ride route A in stead of driving it if I will be safer. 

 

 

My thoughts

 

 

My personal feeling based on years of reading and writing about this is that bicycling is somewhat more "dangerous" than driving. Which makes some sense, since, unlike a car, a bike is inherently unstable at low speeds. And it is really hard to add safety features like seat belts and air bags. Which means that

 

1. While cyclists are less likely to be in crashes with other road users than motorists are, they suffer more "incidents" due to single-bike crashes. Drivers never (Or at least rarely)  fall out of their car, and they don't often lose control if they hit a bad pothole, etc... As a result there are a lot of broken collar-bones and wrists and a lot of abrasions that drivers don't deal with.
2. Cyclists who are involved in crashes have a lower survivability rate than both motorists and pedestrians. Not only are they without any real protection, but they are also up higher (giving them farther to fall), inherently unstable, moving at higher speed than pedestrians and further compromised by the unique geometry of riding (things like clipless pedals and the teeth of the gears). Getting in slightly fewer crashes isn't enough to compensate for the higher likelihood that crashes result in injury or death. 

 

Together  it means that if a person replaces a car ride with a bike ride, I suspect they are something like 5-20% more likely to be injured or killed, depending on where and when they ride. A grade separated bike path during the day might even be safer than driving, but a county highway at night is certainly much more dangerous.

 

But that risk is certainly more than compensated for by the added health benefits of cycling. It's a higher risk, larger reward scenario. But if safety is your main concern, then by all means the bus or rail (or even better, the elevator) is your best option.

Some nits to pick

As it turns out, bikes are the most dangerous way to get around with the exception of motorcycles.

Actually, the cited study also shows "other vehicles" defined as "large truck, motor home, taxi, limousine, hotel/airport shuttle, etc..." as more dangerous than biking. Being 3rd most dangerous out of 6 options isn't exactly "call the neighbors and wake the kids" territory. 

Bike commuters inhale about three times as much air pollution as drivers

On identical routes. If your bike ride takes you on the Captal Crescent or Mount Vernon Trails, that is almost surely untrue - as long as you aren't adding a lot of miles. 

“We think bikers are choosing a healthier option, but it turns out bikers on long, crowded roads are getting exposed to air pollution,” he says.

Bikers are choosing a healthier option AND they are getting exposed to air pollution.

 

Comments

[Brett Young]

I stopped reading the article that Sadie wrote when she said "In addition to accidents, cyclists face another major health risk: Air pollution. Bike commuters inhale about three times as much air pollution as drivers..."

Wow. And to think how unsafe walking is...especially those long distances...think of all of that air pollution that I inhale.

[jeffB]

Well I still think it is a crime that there are days when the air quality is so bad that it is recommended people stay indoors.

Wake up people. We only get 1 planet.

[Crickey7]

Apparently, I'm already dead and don't even know it.

[john]

tl;dr: Bike commuters live longer than auto commuters. The two groups don't always die from exactly the same things.

[SJE]

Cyclists are less risk averse, and that is associated with all sorts of increased risks of physical injury and accidental death from all activities. (e.g. my willingness to try different foods led me to 3 days of GI distress).

But a slightly higher risk tolerance is also associated with lots of positive benefits: health, creativity, entrepreneurship, and greater financial success. It was two bicycle mechanics that flew the first powered plane.

[Zack Rules]

I wish she would have mentioned how NO ONE has died using bikeshare in the US despite 30m+ rides since 2007. There is literally no other transportation mode that is safer.

[Smedley Burkhart]

Mortality is a very limited way of looking at risk-benefit ratio of cycling. I bet a measure of significant health impairment would give a different picture. I also don't think the mortality stats have full applicability to us, Dears. We all hear about the tragic deaths of local bourgeoisie, but the nationwide data capture clueless youth and the involuntarily car-free, as well, both of whom probably contribute significant numbers.

As for exposure to air pollution, that is a risk, not an outcome, and there is no basis on which to assume that any related health outcome would be the same in a sedentary and a fit population.

[DKB]

And are motorcycles the most dangerous way to "get around?" Or are they the most dangerous way to get home from a bar? Or are they the most dangerous vehicle for men between 18 and 29? Or..?