In Defence of the Bicycle

in which I try to prove that cycling and walking are not dangerous

Related articles:
This document started out as a discussion on a local mailing list about bikes and bikes issues in Santa Cruz, CA. In the course of a different thread, I said testily that "cycling is not dangerous." A professional "bike planner" currently working for our local government immediately responded (his text appears in this font). I disagreed strongly with his approach and wrote this lengthy rebuttal (Aug 15 2001).

>> Accident statistics indicate that cycling is, unfortunately, 
>> dangerous,

>>      "On a per-trip basis, walking and cycling are roughly 
>>      three times as dangerous as riding in a car.  In 1995, 
>>      there were 29 ped. fatalities, 26 cyclist fatalities, 
>>      but only 9 car occupant fatalities per 100 million 
>>      person trips.  Walking and cycling appear even more 
>>      dangerous when these fatality rates are calculated on 
>>      the basis of distance traveled.  Per kilometer 
>>      traveled, walking is 36 times more likely to result in 
>>      fatal injury than riding in a car; cycling is 11 times 
>>      more likely to result in death.  In short, the dangers 
>>      of walking and cycling in America are not just 
>>      perceived; they are real."

To my ears, what Heath just posted (about walking and biking being very dangerous) sounds like a traditionalist view of road/traffic safety :-)

For at least 25 years, maybe a bit longer, there has been another, contrarian view -- what we might call the "New School," as opposed to the "Old School" that dates back over 50 years. In the last 5-10 years, as critiques of automobile-dependence have slowly filtered from way-out-in-left-field to a precarious foothold with the most "leading edge" established authorities, the debate between these two viewpoints has intensified.

After a lot of reading & thought, I've become a card-carrying member of the New School :-) So here are a few "New School" thoughts, to counterbalance the establishment view...

Risk per mile is, imco, a deeply biased (and discredited) metric. It has been widely criticized, mostly because it always makes the fastest mode of travel look safest. Americans drive 2.5 trillion miles a year, collectively. Risk per mile is indeed very low, but people drive enough miles -- raise their exposure per annum high enough -- that we kill 35,000 people inside cars every year on our roads.

  • short essay on RPM metric

    Risk per trip is a better metric, but is difficult to calculate because of huge variation in trip length. Risk per hour of exposure is the least biased of all metrics; and as Kevin points out, cycling does not stand out as particularly risky when considered per exposure-hour. It is, however, difficult to get accurate figures on BHT/annum.

    We must also bear in mind that all road risks experienced by Americans are relatively low on whole-pop scales of risk. Let's take a whole-population view of annual risks: there's a 1/600 chance that you will commit suicide, and a 1/11,000 chance that you will die as a car occupant, this year. Your average chance of untimely death from all causes is about 1/115. Your population-based risk of dying while cycling this year is only about 1/130,000.

    Here one must rightly object that only 6 percent of Americans actually cycle regularly -- of course the whole-pop risk looks low! Sadly true, and so we must adjust that risk upward by a factor of about 16 to compare it to the whole-pop risk of death in a car. [This factor of 16 is based on an assumption that almost 100 percent of Americans either ride in, or drive, cars regularly.] If almost every person in the US cycled (as many as now travel by car), then projecting from current risk levels, we would expect the cyclist annual fatality risk to be 16 times higher or 1/8,125 ... in the same ballpark as the 1/11,000 risk from driving or riding in a car: less than 1.5 times as risky.

    But this is misleading. In actuality, the risk of cycling seems (from international comparative studies) to be inversely proportional to the number of cyclists on the road. The more cyclists, the safer it is per cyclist per annum.

    There are various theories about this Smeed's Law effect, but common sense tells us that (a) if more people are on bikes then fewer people are driving, and (b) if more people are on bikes, then bikes are more common, more accepted, expected, and not stigmatised as a weird or inferior transport option. Both of these factors would naturally improve cyclist safety. So if the number of Americans cycling did in fact increase markedly, we would expect the whole-pop risk per cyclist to go down -- the annual risk per cyclist for 16 times as many cyclists as we have today, would be considerably less than 1/8,125 and would look better compared to automobile occupant risk.

    In general, Americans have an exaggerated subjective perception of the risk of walking or cycling, and a muted perception of the risk of driving.

    But the fact is that even with respect to routine risks that we all run (for example, heart attack, cancer, or traffic fatality) there is generally a huge discrepancy between the true magnitude of a risk and the layperson's perception of it. For instance, ... the average American reckons the odds of his or her dying in a car accident this year to be about 1 in 70,000; the real figure is closer to 1 in 7,000.

    (Larry Laudan, The Book of Risks)

    The redefinition of walking and cycling as "dangerous" activities keeps most Americans sedentary (i.e. inactive, travelling only by automobile). I find that by far the most common reason my car-dependent friends give for using their automobiles for trivial trips is simple fear. They are afraid to ride a bike on any public road, or to walk downtown. But ironically this sedentary, car-dependent lifestyle contributes significantly to health problems which pose a far greater risk than the risk of injury or death while travelling.

    Your chance of a heart attack this year, if you are an American over the age of 35, is one in 77! Your chance of being killed by a car while walking is one in 40,000. You have a pretty good chance of diminishing the first, major risk, by undertaking the second, relatively minor risk. Furthermore, if more people walked instead of driving, the health benefit of walking for each individual would remain just as high, but the risk of walking for all would diminish. This makes alternative modes of transit an inverse "tragedy of the commons" -- in this case smart for one is smart for all.

    The British Medical Journal a couple of years back published an excellent paper on cycling, which attempted to balance road risk against health benefits; this was a followup to earlier studies which found that regular cyclists tend to live several years longer, and enjoy better health, than sedentary persons. The researchers found that, approximating as best they could on the basis of national health statistics, road fatalities and injuries, etc, the benefits of cycling outweighed the risks by 20 to 1. In other words, you are twenty times more likely to lengthen your life than shorten it by cycling regularly.

    >> Perpetuating the myth that cycling here does not involve an 
    >> element of risk does a disservice those who would like to 
    >> see cycling made safer. 

    I doubt any reasonable person has ever claimed that cycling involves no element of risk :-)

    The only way to avoid all risk is to be dead already. Every activity we undertake involves risk. Refraining from all activity also involves risk. So there is no way to be alive and not take risks. My point is that cycling is not dangerous as compared to other activities that we do not designate as particularly risky, i.e. not as dangerous as it is painted -- and that cycling and walking also facilitate risk reduction (by improving general health).

    Aside from exaggerating road risk, we also (imho) do ourselves a great disservice by displacing it. Almost all cyclists and pedestrians who are seriously injured or killed on our roads are struck by cars or other motor vehicles. Repeat, most serious injuries and deaths experienced by cyclists and pedestrians on public streets are a result of being struck, dragged, or run over by a motor vehicle. The kinds of accidents cyclists and peds get into without the "help" of motor vehicles are generally trivial, not life-threatening, and not productive of major injury or disability.

    Therefore the real danger is not inherent in walking or cycling. The WHO once estimated that cycling in the absence of motor vehicles is about 500 times safer than driving in the absence of cyclists :-)

    The danger (such as it is) is introduced by an excessive number of heavy vehicles in rapid motion. By concentrating exclusively on the pedestrian or cyclist as the person "taking a risk" or indulging in unsafe behaviour, we render invisible the person who, by choosing to drive, actually creates that danger and inflicts it on others.

    Even those behaviours which Kevin rightly describes as risk-enhancing for cyclists are far less hazardous in a less carcentric country:

    A few summers ago I was in Amsterdam sitting in a street-side cafe at midnight. The place was swarming with bicycles - almost all of them unlit. And the fatal accident rate for cyclists there is v low compared to Britain. Says something I suspect about the expectations of all those involved...

    (J Adams of the Institute for Policy Studies, London; personal correspondence)

    We also reduce the alertness and conscientiousness of drivers, by making people feel "safe" and relaxed about driving, rather than acutely conscious of doing something risky. People feel so safe and confident in their cars that, as we all know, they talk on cell phones, read maps or even newspapers, shave, apply make-up, drink, eat, change CDs, conduct intense arguments with their passengers, etc. -- all while driving. Most people do not feel that driving is a risky activity requiring constant attention and concentration; it is a normal, "safe", everyday activity to be undertaken casually.

    By contrast, we (our educators, safety authorities, et al.) lecture and berate pedestrians and cyclists about the need to take care, be hyperalert, never make a false move. But the danger (such as it is) springs from quite another source and thus these lectures have little effect -- except to scare people out of cycling or walking, and to inspire them to prohibit their children from cycling or walking [see recent CDC hand-waving about physical unfitness in American children, and the long-term health costs CDC expects as a result]. Parents then drive their children everywhere -- which compounds the original problem by putting yet more cars on the road.

    The same tragic paradox is manifest in the advice traditionally given by road safety authorities to drivers: always choose the heaviest vehicle you can afford, to keep your family safe. No mention is made of the increase in your risk of killing or injuring the occupants of another vehicle in the event of a crash, when your vehicle is significantly heavier. [cf. NYT article earlier this summer on SUVs and fatality risk, particularly the Ford Explorer] The logical outcome of the policy "drive a heavier vehicle to be safe" is an arms race which pushes the average vehicle weight of the national fleet steadily upwards, thus increasing total road danger. Smart for one, dumb for all.

    By contrast, your risk of injuring or killing another human being if you walk or cycle is a very, very small number -- approx. zero for a pedestrian and (just off the top of my head) something less than one in 1.5 million (fatality) for a cyclist in the US, annually.

    If we assume that 60 percent of the population of any motorised nation are qualified drivers, then in a population of 286 million, there are about 171.5 million drivers. About 50,000 people are killed by motor vehicles every year (all roles: occupants, drivers, pedestrians, cyclists) so as a driver, your share of the whole-pop risk of killing someone (yourself included) while driving is about one in 3400. [This stat actually gets messy because of the infliction of multiple deaths by single at-fault drivers, but we're in the ballpark]. It is slightly higher than your own risk of dying in a car yourself, mostly because of those 5000 pedestrians killed by cars each year. The less-than-800 dead cyclists don't influence the stats much, I am happy to say.

    I am not counting the approx 50,000 other people who are said by CDC and other health agencies to die prematurely due to diseases caused by air pollution -- a death toll for which the average cyclist's or pedestrian's share of culpability is not more than half that of an habitual driver's, and possibly less.

    When I choose to bike or walk instead of driving, I may incur (for myself alone, and through no fault of my own) a higher risk per trip than a driver. We could argue all day about risk per hour, risk per trip, etc. and exactly how much (if any) additional risk I incur. But there is absolutely no question that the risk I inflict on others is thousands of times smaller than it would have been if I drove. The net effect on the community is beneficial. So, am I doing something dangerous? I think not. Is a car driver doing something safe? I think not. The problem is that our definitions of "safe" and "dangerous" are skewed.

    Cycling and walking are far, far less likely to "result in death" (of someone else) than driving a car. The reduction in the risk of hurting or killing another person far outweighs the increase in risk to oneself. Net positive.

    So . . . the traditionalist school of road safety finds cycling and walking to be "very dangerous", despite all the above. This is because it defines "safety" in terms of risk only to the vehicle operator or passenger (traveller), and not in terms of the net safety of the whole community of road users. The contrarian "new school" point of view is that safety must be defined in terms of risk-inflicted as well as risk-run -- and furthermore, that more onus should be placed on risk inflicted on others than on risk freely chosen for oneself.

    For a more complete exposition (with some repetition of points made above, plus additional text about risk, risk measurement, etc) see


    and (draft in progress)

    plus a pile of related publications by e.g. Whitelegg, Hillman, Adams, Franklin, Davis, Fisher, Smith and Komanoff...

    Please note that I am not saying (as e.g. Forester sometimes says) that there's no need to make our streets safer and more pleasant for peds and bikes. Things are not OK the way they are. The need is there. All that noise, stink, rudeness, physical obstruction and ugliness are discouraging to cyclists and walkers; but so is our exaggerated perception of a real (but statistically unimpressive) danger.

    By continuing to redefine cycling and walking on public streets as highly dangerous activities, we either discourage the public from cycling and walking at all -- or we transform cycling and walking into purely recreational activities, which are only pursued in special recreational settings strictly segregated from the quotidian life of the community. Either of these outcomes keeps us locked into automobile-dependence. People start driving extra miles just to get to places where they want to walk or bike for sport/recreation! We also continue in a state of denial about the risks created and imposed by automobile-dependence, and therefore we remain unable to solve that problem.

    For all these reasons I object emphatically to this redefinition of harmless, low-impact, sustainable, healthy modes of travel like walking and cycling as "dangerous" activities. I object to perpetuating the myth that driving is "safer" than walking or cycling, when driving is the primary generator of road risk. Let's be honest about where the danger is coming from. Let's start admitting that our roads are dangerous only to the extent that we have made them so, by allowing (wholeheartedly encouraging) the complete dominance of private automobile transport. There is nothing natural, necessary, inherent or inevitable about this.

    Let's start praising -- not chiding -- pedestrians and cyclists for increasing safety by their choice of low-impact transport. Let's stop saying that people are at risk because they are cycling or walking. They are at risk only because too many other people are driving; because we've engineered our entire urban landscape for the convenience of the most risky, i.e. risk-inflicting, transport mode.

    And when (mercifully rare) tragedy does occur, and a ped or a bike dies after being struck by a car -- instead of saying that this person would still be alive if only he/she had driven instead of walking or cycling, let's reflect on the fact that he/she would still be alive if only the driver had walked or biked that day instead of driving.
    De Clarke