Biking is safer than it used to be.

Biking is increasing in Cascadia's cities; cycling crashes are not. As a result, the crash rate for cyclists has declined, by 70 percent in the Rose City, according to the City of Portland. In Vancouver, B.C., during the period in which the number of cycling trips has almost tripled, the number of insurance claims involving bicycle collisions hasn't budged, according to the Vancouver Courier (sorry, the link is no longer available).

Cycling is safer than you might believe.

Any activity that allows you to travel fast, unshielded, and unrestrained involves risk -- whether it's cycling, or skiing, or base jumping (insane video). Cycling involves the added risks that you're balancing on two wheels and that you're surrounded by moving one- and two-ton steel boxes. If you're trying to avoid getting hit by such objects, being on a small, nimble vehicle is an advantage, as Peter discovered when he was cut off. (Former University of Washington professor William Moritz conducted surveys a decade ago showing that more than 80 percent of bike wrecks -- generally the less serious ones -- involve cyclists falling or colliding with things other than a moving car or truck.)

But if you actually are hit (and car-bike collisions are usually the dangerous wrecks), you'll do far better strapped into a steel case of your own than if you're astride a two wheeler. The bigger the case and the more restraints and cushioning you've got, the better you'll do. So getting hit on a bike is worse than getting hit in a car, which is worse than getting hit in a bus, which is worse than getting hit in a train or a Bradley Fighting Vehicle.

In fact, the best published estimates I've found -- developed by Rutgers University researchers John Pucher and Lewis Dijkstra -- suggest that per trip, bike riders face about three times as much risk of dying as car and light truck riders. Because car trips tend to be longer than bike trips, Pucher and Dijkstra estimate that the safety gap stretches to tenfold when it's calculated per mile traveled.

That's a substantial gap, if Pucher and Dijkstra are right. But how big is the risk, really? In the United States, for every billion kilometers of cycling, they say, roughly 100 bikers die from collisions. For every billion kilometers of driving, roughly 10 drivers and passengers die from collisions. From 1999 to 2004, in the entire United States with its approximately 300 million residents, an average of 784 people died each year in bike accidents. That's a consequential number but it's no pandemic -- nothing like the more-than-40,000 deaths from auto accidents each year. It's reason for care but not for alarm.

Four other pieces of information put these figures in context. First, if the danger of cycling seems excessive, consider that riding a bus or train is 10 times safer than riding in a car, per mile, according to the same researchers. Many people won't bike because it's "too dangerous," but not many people refuse to drive because transit is so much safer. Why? As I've said, we're not entirely rational about transportation decisions.

Second, the same published estimates indicate you're at much greater risk of getting hit by a car when you're walking than when you're cycling. Per mile traveled, according to Pucher and Dijkstra, more than three times as many pedestrians die from auto collisions as do cyclists. Yet few people think walking is too perilous to attempt. (Ditto re: rationality.)

Third, Pucher and Dijsktra may be wrong. The statistical challenge that all safety analysts face is that no one really knows how much cycling -- or walking -- people do. Estimates vary widely. Pucher and Dijkstra accept a low figure for total cycling to calculate accident risk. Others use higher figures for cycling, which makes crashes seem less common. In the early 1990s, for example, Failure Analysis Associates (since renamed Exponent), one of the world's leading engineering firms in the specialty field of quantifying risk exposure and preventing mechanical failure, estimated that riding in a car for an hour is almost twice as likely to kill you as is riding a bike for an hour. Repeat: this credible source suggests that biking is not more dangerous than driving but is, in fact, half as dangerous. Unfortunately, the analysis was proprietary. Only one summary table (see below) is in the public domain. The engineering journal Design News published it with little comment in 1993 in an article on a different subject. (I've asked Exponent for supporting documentation but have yet to hear back, probably because the estimates are so old. I'll update this if I learn more.) Several bicycling advocates tout this table, and one has even demonstrated its quantitative plausibility. Still, as of now, Pucher and Dijsktra's estimates are the only ones published in a peer-reviewed journal, so I'll assume they're about right.

Fourth, because of the widespread perception that cycling is dangerous, the existing population of cyclists may be disproportionately made up of risk-takers. If everyone thinks biking is unsafe, the people who do it will be the ones who don't mind danger. And such people are more likely to get hurt in just about any activity. In his 2004 book The Art of Urban Cycling, Robert Hurst cites evidence that as many as half of car-bike crashes are the cyclist's fault: the cyclist ran a stop sign, made an illegal turn, rode against traffic, or otherwise broke the law.

(Aside: a smattering of bike riders clearly seek out risk intentionally. They're risk junkies, shown in the following video "drag racing" through New York).

What this means is that if you're a cautious, law-abiding, risk-averse cyclist, biking is far safer than you'd think from the aggregate statistics, which are inflated by the proliferation of two-wheeling daredevils.

Put all these considerations together and it looks like the added increment of crash danger you put yourself in from biking, rather than driving, is small, if it exists at all. Furthermore, if you care about not imperiling others -- assuming you want to avoid both dying and killing in a collision -- then cycling looks substantially safer than driving, because bikers almost never kill or injure others. But even assuming you don't care about anyone but yourself, cycling is still the healthy choice, because crash danger isn't the end of the story.

Biking's health benefits massively outweigh its health risks.

Cycling is the kind of low-impact, moderate exercise that humans need in abundance in order to enjoy vigorous, healthful lives. One study (hat tip to Todd Litman) followed almost 30,000 Danes, monitoring their physical activity and health. Lars Andersen and his co-authors concluded, "Even after adjusting for other risk factors, including leisure time physical activity, those who did not cycle to work experienced a 39 percent higher mortality rate than those who did." In other words, nonbikers -- even if they were active in sports -- died 40 percent more often than bikers.

Similarly, Pedalling Health, an Australian study published in 1996, concluded that an hour of biking a day -- normal for a regular bike commuter -- prevents four times as much heart attack risk as it adds in collision risk. The iconoclastic British transport researcher Mayer Hillman did a study for the British Medical Association in 1992 (not online but summarized here and here) reportedly showing that for every year of life lost to a bike crash, twenty years of life are gained from stress reduction, greater cardiovascular fitness, and improved mental health. As I've noted, the time you spend in moderate exercise is added to your life, with interest.

Cycling is not as safe as it should be.

Still, in the world's centers of Bicycle Respect, cycling is radically safer. Pucher and Dijkstra wrote in the American Journal of Public Health (PDF), in 2003, "per trip cycled, American bicyclists are twice as likely to get killed as German cyclists and over three times as likely to get killed as Dutch cyclists." Per kilometer of travel, the gap is larger: Dutch cyclists are more than ten times safer than their American counterparts. European safety records are improving faster, too. In Germany, for example, collision deaths per bicycle trip have fallen by more than 80 percent since 1975, according to Pucher and Dijkstra. (Of course, we should use caution with all these figures, because, like previous ones, they all depend on estimates of how much biking people do.)

Making cycling safer is a main chance for healthy, lasting prosperity.

I began this by wondering about my son Peter, but it turns out that bike safety isn't just an issue for worried parents. A few years ago, when the Puget Sound Regional Council asked cyclists why they did not commute by bike more often, the leading answer by far was "unsafe routes." This finding was not unusual. Pucher and Dijkstra, writing in Transportation Quarterly this time, noted, "Almost every survey finds that the perceived danger of cycling ... is one of the major deterrents to increased bicycle use in the U.S." People are afraid of traffic.

Making cycling safer, therefore, may unleash more two-wheeled travel more than any other thing that communities can do, with huge benefits in stemming obesity, oil imports, and climate disruption.

The keys to cycling safety in Europe are facilities, traffic laws and enforcement, education, and numbers:

1. Good cycling facilities. Bikeways, bike boulevards, traffic calming, blue lanes, and cycle signals save lives. In Copenhagen, for example, major intersections painted with "blue lanes" to mark bicycle routes have seen a 40 percent drop in deaths and injuries to cyclists. In Denmark, Germany, the Netherlands, and the United Kingdom, traffic-calmed neighborhoods -- those where streets re-engineered with curb bubbles, traffic circles, and the like slow car traffic and shield nonmotorized travelers -- have seen half of traffic injuries disappear.
2. Bike- and pedestrian-oriented traffic laws save lives, too. For example, one key difference that helps make northern European cyclists safe is low speed limits. In Germany and the Netherlands, speed limits in residential neighborhoods are commonly under 20 miles per hour, while arterial speeds are typically limited to about 30 miles per hour. Slower driving means fewer -- and softer -- impacts: 95 percent of people hit by a car at 20 miles per hour survive; just 15 percent survive at 40 mph, as we noted in Cascadia Scorecard 2006 (see page 52, login or registration may be required).
3. Educating drivers and cyclists also boosts safety. German and other northern European drivers licenses are much harder to get: you must be 18 and have completed a rigorous and expensive private training course. But traffic education is not just for drivers. German and Dutch schools, for example, provide comprehensive cycling and walking instruction to all schoolchildren by the time they're 10 years old, as detailed by Pucher and Dijkstra. German third- and fourth-graders take bicycling classes, at the end of which they demonstrate their skills to traffic police on special courses that simulate local streets.
   
   In Cascadia, school-based cycling and pedestrian education is mostly limited to the occasional bike-to-school day. Notable exceptions are found in Corvallis and Eugene, Oregon. Eugene runs cycling school buses: parent volunteers lead groups of bike-riding students to school together, following routes that collect more students as they go. And Corvallis provides week-long bicycle training for all fifth graders.
   
   My ultimate hope is that Cascadian communities will replace "driver's ed" with mobility education -- training young people to safely navigate their communities by a variety of means, from automobile to bicycle to foot to transit. (The just-formed champion for this bold and necessary idea is the new Seattle-based Mobility Education Foundation.)
4. Safety in numbers. As we noted before, the more cyclists and pedestrians on the streets, the safer they become. Health consultant Peter Jacobsen of Sacramento, California argues in the journal Injury Prevention that it's because drivers become more attentive as cyclists proliferate. It could also be that cyclists become more law-abiding as they proliferate, if the risk-takers already on bikes are being joined by large numbers of risk-avoiders. Either way, the beauty of this finding is that safety improvements in facilities, education, and law enforcement will induce additional cycling, which will bring further safety improvements through numbers -- a kind of virtuous cycle (sorry).

Back to where I began: is cycling safe enough for my son Peter? Easily. The modest risks are swamped by the benefits. Still, it's not as safe as it should be. All by myself, I cannot give Peter the safety levels of Germany or the Netherlands. I can't personally install city-wide bike facilities, pass new traffic laws, and provide comprehensive mobility education to all. But I can carefully choose his routes with him, teach him to ride legally and cautiously, and provide ongoing education about how to get around safely.

Beyond that, I just have to remember that what's really dangerous isn't biking (or walking), it's sitting around. Not pedaling can kill you.