More Evidence That Helmet Laws Don’t Work

There was a correlation between living in an area with high cycling rates and low levels of hospitalization. Graph: University of British Columbia
Living in an area with high cycling rates is linked to lower levels of hospitalization for bicyclists. There is no similar link for helmet laws. Graph: University of British Columbia

If you want to increase cycling safety in your city, drop the helmet law and focus on getting more people– particularly women — on bikes, with street designs that offer separation from vehicle traffic.

That’s the finding of a new study by researchers at the University of British Columbia [PDF] evaluating safety outcomes for cyclists across Canadian provinces and territories.

Lead author Kay Teschke and a team of researchers looked at cyclist injuries requiring hospitalization in 10 Canadian provinces and three territories between 2006 and 2012. They checked to see if hospitalization rates were linked in any way to helmet laws and cycling rates, and they checked for variations in hospitalization rates by sex and age.

Helmet laws were found to have no relationship to hospitalization rates. That was true even though self-reported helmet use is higher in areas of Canada that mandate it (67 percent) than in areas that don’t (39 percent).

But having a higher rate of cycling in one’s community does seem to have an impact on safety. Using Canadian government data on cycling activity, researchers found that men and woman were both less likely to be injured while biking in communities where more people bike.

It’s not the first time this effect, sometimes called “safety in numbers,” has been observed, though the researchers wrote in their summary of the study, “the explanation could also be ‘numbers in safety’ as safer bicycling infrastructure has been shown to attract more people to cycle.”

The researchers also found that women were much less likely to be hospitalized for a cycling injury than men, echoing the results of other studies examining traffic collisions of all kinds. The authors attribute this to a “lower propensity for risk taking” among women.

“Transportation and health policymakers who aim to reduce bicycling injury rates in the population should focus on factors related to increased cycling mode share and female cycling choices,” the authors conclude.

31 thoughts on More Evidence That Helmet Laws Don’t Work

  1. I literally posited this connection between policies that discourage cycling & the safety in numbers effect, during a transportation planning class at McGill in 2010 & was browbeaten by the professor…sigh.

    The perception of safety is very different than actual safety, but safety in numbers is stronger than both. We need to build safety improvements that encourage cycling, not require safety precautions that discourage it.

  2. Any analysis that purports to debunk helmet-wearing by the use of aggregates or overall rates of injury and/or hospitalisation misses the point.

    Of course most bike rides don’t wind up requiring a helmet. And of course most injuries suffered on a bicycle have no relation to whether the bicyclist was or was not wearing a helmet.

    But helmet use is not about aggregates or averages or even likelihoods. Helmet use is about that one case where the helmet comes in handy; it is about that one time where the helmet makes the difference between walking away from a crash and suffering a life-altering permanent brain injury.

    That sort of crash is extremely unlikely. It will probably never happen to you or to anyone whom you know. But it will certainly happen to someone eventually.

    Rational risk assessment is done by multiplying likelihood by harm. Wearing a helmet is thus a reasonable precaution against this sort of unlikely yet potentially devastating crash. For this reason, bike advocates are morally obligated to model and to defend the practice of wearing helmets.

  3. The problem with your reasoning is head injury is actually more likely while walking. Therefore, we should advocate pedestrian helmets even more strongly than bike helmets to protect against this highly unlikely but severely debilitating outcome. In fact, we should advocate helmets in automobiles as well for exactly the same reasons.

    The reality here is when the likelihood of something is extremely low, it makes little sense to protect against it unless there are zero downsides to that protection. Unfortunately, bike helmets, really any helmets, have far from zero downsides. They actually have enough downsides, such that they discourage cycling enough to result in a real increase in heart disease from inactivity. This kills far more people than helmets might save.

    Bottom line—helmet use should remain an individual choice, with cycling organizations (and legislators) taking a neutral stance. That means towards everyone, including children. I can say with some certainty had child helmet laws been in effect when I was young, I would have had zero interest in learning to ride a bike. The end result of that would most likely mean I would now be a 300 pound poster child for heart disease.

  4. The study does not seek to “debunk helmet-wearing”, it seeks to see if helmet LAWS increase safety. The data they use is not perfect of course as a way to measure this. But hospitalization is a decent way to measure if something had any effect on the kind of safety that helmets are supposed to impart.

    There is a separate, but valid debate to be had about whether wearing a helmet generally increases safety. There’s some evidence that those wearing helmets ‘risk compensate’ and ride more dangerously while wearing a helmet. There’s also some evidence that certain head and neck injuries can be made worse by a helmet. Obviously there are also plenty of ways in which a helmet could prevent serious injury. But overall that questions isn’t as clear as you would think.

    But what is becoming increasingly clear is that helmet laws are at best unhelpful and at worst actually putting riders in danger by discouraging cycling. For this and several other reasons, bike advocates are actually NOT morally obligated to model and to defend the practice of wearing helmets.

  5. When I was hit by a car while riding and wearing a helmet, I still went to the hospital- for my knee. But my neck and head and brain were fine. Why are you showing us a chart of hospitalizations irrespective of body part?

  6. Because they are trying to correlate helmet laws to overall safety. They are not trying to connect wearing a helmet to severity of certain types of injury.

  7. With current road speeds, the number of hospitalizations tracks the number of collisions. That seems self-evident. Helmets don’t reduce the number of crashes, so they don’t reduce the number of hospitalizations. I don’t know that that will compel the general public to believe that they do not serve an important protective function.

  8. This article is about macro-level safety of helmet culture. However, even at the micro-level (aka “selfish” individual level) there are reasons to be skeptical about bicycle helmets:

    –> bicycle helmets are no motorcycle helmets. In other words, their safety benefits to people in biking situations is unfortunately quite dubious:

    –> while the idea of helmets was a well-intentioned concept, the law of unintended consequences apples–studies show that many people tend to unintentionally bike at least a bit more recklessly when helmeted. Some a lot more so. Drivers also tend to drive more recklessly around individuals with helmets on.

    –> since helmet culture discourages biking and strength in numbers is the single biggest safety factor in biking, it’s in your selfish interest to be surrounded by people on bikes–regardless of what gear they have on. Unfortunately, helmet culture significantly blunts biking numbers.

  9. By the same logic, we should have mandatory car helmet laws.

    And people should definitely wear helmets when taking a shower. Lots of slips and falls and head injuries there.

  10. Speaking of Canada and helmet laws, the other major problem is the extreme inequity in terms of selective enforcement (this is also very true if not more so in the US).

    This can literally impact your safety/dignity/well-being due to police harassment if you’re biking helmetless while brown, as Vancouverite Andishae Akhavan discovered a couple years ago while biking helmetless in Vancouver:

    BC is one of the few places in the world with an all-ages mandatory bicycle helmet law and it gives police the “excuse” to disproportionately pull over minorities for the most minor of infractions.

    Meanwhile, Vancouver police almost never ticket white European-tourist-looking types biking helmetless around but the stats show that if you bike helmetless while brown you’re really asking for it. Shameful.

  11. Another long term, large scale, reliable study shows no benefit from helmet wearing by cyclists, so why do so many people believe the short term, small scale, unreliable studies? Whatever, the fact that nowhere with a helmet law can show any reduction in risk to cyclists, despite more than twenty years of data, is proof enough.

    The only two detectable effects of helmet laws and propaganda are to reduce the number of cyclists and to make obscene profits for the people making and selling helmets. Because the people deterred from cycling lose the overwhelming health benefits (regular cyclists live longer, are healthier, fitter and slimmer) they get fat, sick and are a massive burden to health services.

    It’s time the helmet proponents looked at the actual effects of helmet propaganda and laws, not the ones they fondly imagine will happen. Check out the evidence at

  12. I find this article misleading, of course bicyclist infrastructure is going to have an impact on injuries and accidents… But helmets still help prevent severe injuries. You only have one brain. However, if there is a helmet law it needs to be enforced equitably and not unfairly target people of color. That’s wrong.

  13. The need to wear a helmet can “discourage cycling” only in people who are more concerned with their hairstyles than with their skulls and brains.

    I am bald now; bu I had wonderful long hair as a kid. So I know what it means to fuss over how your hair looks. But I also know how to order my priorities.

    Anyone who values avoiding hat-flattened hair above all else is simply too shallow to be reached. Sane people protect their heads.

  14. Most bicycle helmets are good for about banging your head on a low branch as you pass by. For anything else, your head bounces around in the cavity too much for it to be of much use. It’s the brain bouncing around in the skull that is most damaging-usually. External impacts are of some issue…but not usually.

    Want to protect your noggin? Wear a 3/4 shell helmet. Want to really protect? Wear a full fact mountain bike helmet.

    Or…wear nothing and actually enjoy the ride.

  15. The legal requirement to wear a helmet can discourage cycling by creating the impression that it is a highly risky behavior (helmets are generally only required in activities where there is a “high perceived risk” of head injury). As has been pointed out elsewhere, we do many daily activities that have as high or higher risk of head injury where we never even discuss the idea of people wearing helmets. They also add expense, inconvenience, and discomfort.

    Are any of those strong reasons why any individual should not wear a helmet? Generally no. But they are good reasons why helmet laws are not good policy. The research here and elsewhere shows that the most effective ways of making cyclists safer is to have more cyclists and to have better infrastructure. Helmet laws discourage cycling regardless of whether or not they improve the personal safety of those wearing them.

    Some things are more complicated than “helmets good” or “helmets bad”. Wearing a helmet probably reduces your chances of serious head injury. Helmet LAWS get people killed. Whining about how people who try to engage with this in a complex, reasoned based way built on evidence are somehow insane or shallow doesn’t help your case.

  16. …why do so many people believe the short term, small scale, unreliable studies?…

    ———- intuition. very few people go after real facts and statistics. the best source (although a few years old, now ) is:

    ——— the author is still around, so if you have questions, contact him . . .

  17. Actually, bicycle helmets would be even more useful for pedestrians, who face a greater risk of head injury and for whom the helmets would actually ironically work better:

    Of course, it’s still your personal choice whether to do one or the other, but it’s more than a little facile to brand non-wearing as vanity.

    After all, unless you’re being logically consistent by also wearing your helmet anytime you also walk somewhere (seeing as head injuries occur oftener amongst pedestrians and these kinds of helmets could help at least a percentage of these injuries), your reasoning is also much less likely fact-based than emotion based.

  18. The problem is that bicycle helmets’ safety record seems to show they don’t prevent all that many injuries in the first place. After all, even bicycle helmet manufacturers hedge their bets by stating they’re of use only in 1) non-potentially fatal crashes with 2) non-vehicles at 3) slow speeds 4) to the crown of the head (which is not where most people hit the ground upon falling off a bike).

    Jurisdictions which have mandated helmet usage have unfortunately tended to see head injuries increase per capita after the implementation of the laws, which is a major reason they haven’t become widespread. For example, one reason a proposed helmet law in California was rejected last year was due to projections showing it would likely *increase* healthcare spending, as has generally happened in other places it’s been implemented.

    And, as you point out, targeting POC is a major concern. As Vancouver’s example shows it’s virtually impossible it’d be doled out equitably. Do we even need to pretend to guess which of the following people is more likely to be apprehended by police for not wearing a helmet?

  19. The police would probably stop the girl also—to try and hit on her. Not wearing a helmet would give them a ready excuse to do so.

  20. “Anyone who values avoiding hat-flattened hair above all
    else is simply too shallow to be reached.”

    Why is the helmet hair myth rolled out every time as a reason that people don’t wear helmets? I’ve never heard anyone say that, so it’s pretty well an invention and doesn’t say much for your depth.

    “Sane people protect their heads.”

    Sane people read the evidence, they don’t just believe unverified assertions. You could do the same at

  21. “Anyone who values avoiding hat-flattened hair above all else is simply
    too shallow to be reached. Sane people protect their heads.”

    Hear hear! This is why we should require drivers to wear helmets. Thousands of lives would be saved given that head trauma is a common cause of death in auto collisions.

  22. Anecdotes are not data, but for what it’s worth when I was 15 I locked my front brake, flipped the bike over, and landed directly on my head. Helmet smashed apart around my head (still an odd sensation to recall) but my head was OK, albeit bruised.

    They do help at least some people some of the time. I don’t support helmet laws, but I do support helmet use.

  23. “But having a higher rate of cycling in one’s community does seem to have an impact on safety. Using Canadian government data on cycling activity, researchers found that men and woman were both less likely to be injured while biking in communities where more people bike.”

    What is causing what here? Do the cities that have higher rates of ridership have better designed streets? If so, that could be the reason why cities with higher rates of cycling have better safety records. “Safety in numbers” does make sense as a theory, but from what I know that hasn’t necessarily been proven. Determining causal relationships is not necessarily straightforward.

  24. Are pedestrian advocates also “morally obligated to model and defend the practice of wearing helmets”? People have died while walking in parks and being struck by falling tree branches. If, to you, it really is about “that one time where the helmet makes the difference between walking away from a crash…”, then you are morally obligated to where a helmet 24-7, even when you sleep, because there may be that one time where a ceiling collapses from an earthquake or falling tree and strikes one on their head. That “sort of crash is extremely unlikely…but it will certainly happen to someone eventually.”

  25. No one is arguing that helmets themselves serve no protective function at all. But the article summarizes yet another study that once again shows that helmet laws ultimately are not linked to any manner of safety for bicyclists, which is usually the overwhelmingly prominent reason advanced for mandating them in the first place. As such, if they’re not increasing overall safety, spending time and money having a law on the books, especially one that has the propensity for disparate enforcement, is not at all good.

  26. I wear a full helmet every time I ride. Better safe than sorry.

    Though, enforcing a law for it is a bit much.

  27. People aren’t riding around at 10-20 mph. They’re walking.

    Speed is the important factor here. Taking a drop at 25 mph is a lot different than tripping at 3 mph. The same is true for motorbikes.

    It’s common sense to wear a helmet. If you don’t wear one, no matter how you justify it you’re a bad model. You can argue that it’s not fashionable, or that it’s an inconvenience, but at the end it lives are saved.

  28. Pedestrians are hit by drivers every day. You can argue that it’s not fashionable for pedestrians to wear helmets, or that it’s an inconvenience, but at the end lives are saved. For that matter, many people driving die of head injuries, and they are driving at 25-65+ mph. Why shouldn’t they wear helmets? Why shouldn’t everyone wear helmets at all times? Why not? Lives would be saved!

  29. This is a good question, and I suspect a lot of this is chicken-egg. Pure causality may be difficult to prove. Perhaps fruitful, however, are some control cases:

    –> Cities such as NYC, DC and Long Beach and others that have done Before/After studies on protected bike lanes have generally found that in a statistically significant way ridership increases above the general background rate:

    (Cynics occasionally posit that rather than encouraging new bike trips these protected stretches just siphon off existing trips from parallel alternative routes. However, even in the unlikely event that that 100% explained the large increases I’d say that’s still a huge win for these types of treatments as it is literally people voting for what they prefer with their wheels).

    And that injuries decline on these stretches more dramatically than elsewhere:

    –> It’s not uncommon to see raw number of injuries decline even as ridership goes up (which means that the per-capita injury rate is obviously even lower):

    Again, while causality may be tough to fully prove it is worth mentioning that these types of 1) modeshare increases coupled with 2) injury decreases to this degree are rarely if ever seen with unmodified conventional roads.

    Of course a lot of factors go into these individual before/after studies, but whether it’s true causation or not the strong correlation between increased ridership and decreased injury definitely makes a strong case for policy favoring such treatments (with detailed followup studies, of course) as an integral factor in making a city friendlier to more bike trips (better infrastructure, bikeshare, secure/plentiful bike parking, etc.).

  30. That’s the finding of a new study by researchers at the University of British Columbia evaluating safety outcomes for cyclists across Canadian provinces and territories. Lead author Kay Teschke and a team of researchers looked at cyclist injuries requiring hospitalization in 10 Canadian provinces and three territories between 2006 and 2012. They checked to see if hospitalization rates were linked in any way to helmet laws and cycling rates, and they checked for variations in hospitalization rates by sex and age. Helmet laws were found to have no relationship to hospitalization rates. For more details Snowboarding Helmet visit here.

Leave a Reply

Your email address will not be published. Required fields are marked *


WaPo Is Wrong: Head Injuries Are Down, Not Up, in Bike-Share Cities

A Washington Post headline proclaimed today that cyclist head injuries have increased in cities with bike-share systems, based on a study published in the American Journal of Public Health. But University of British Columbia public health professor Kay Ann Teschke is challenging that conclusion, pointing out that the data cited by the WaPo actually leads to […]