What are we Arguing About?

Do Helmets Work? Is There a "Helmet Effect"?

Not much would change if everyone wore helmets. About one to four cyclists get killed by cars in Austin each year. Since many of these cyclists wear helmets anyway, and since cyclists often die when struck by motorists even when they are wearing helmets, it would take about eleven years before the helmet ordinance might save even a single life. In the meantime, many more motorists die in car crashes. The City could save more lives by requiring all motorists to wear helmets. Also, a similar number of pedestrians get killed by cars each year (compared to the number of cyclists killed), but nobody is suggesting that pedestrians be forced to wear helmets.

Finally, many of those cyclists who are killed WERE WEARING HELMETS ANYWAY. It's lunacy to think that a flimsy piece of styrofoam around your head is going to make you invincible against a huge mass of metal moving at ridiculous speeds. REAL bike safety is about not getting hit in the first place, and along those lines, the City won't even get cars out of our bike lanes.

--Michael Bluejay, frustrated cycle activist in Austin TX

I called a local (Massachusetts) Suzuki dealer, and told the salesman I was a first-time buyer looking for something cheaper than the standard $15,000 Harley. He said I could buy the GSXR 1300 for only $10,500, a bike that could hit speeds in excess of 160 miles per hour. He recommended that I wear a helmet, even in non-helmet-law states. Imagine: a novice on a 160-mph bike wearing a plastic hat that will reduce any impact by 14 mph. It's like having sex with King Kong, but bringing a condom for safety's sake.

[...] Why the enthusiasm for helmets? Mike Osborn, chairman of the political action committee of California ABATE, says insurance companies are big supporters of helmet laws, citing the "public burden" argument. That is, reckless bikers sans helmets are raising everyone's car insurance rates by running headlong into plate-glass windows and the like, sustaining expensive head injuries.

Actually, it's true that bikers indirectly jack up the rates of car drivers, but not for the reason you might think. Car drivers plow over bikers at an alarming rate. According to the Second International Congress on Automobile Safety, the car driver is at fault in more than 70%of all car/motorcycle collisions. A typical accident occurs when a motorist illegally makes a left turn into the path of an oncoming motorcycle, turning the biker into an unwitting hood ornament. In such cases, juries tend to award substantial damages to the injured biker. Car insurance premiums go up.

Osborn sees a hidden agenda. "They [the insurance companies] want to get us off the road." Fewer bikes means fewer claims against car drivers. Helmet laws do accomplish that goal, as evidenced by falling motorcycle registrations in helmet-law states. It is interesting to note that carriers of motorcycle insurance do not complain about their clients. Motorcycle liability insurance remains cheap. Osborn pays only $125 per year for property damage and personal injury liability because motorcycles cause little damage to others.

Forbes FYI
'The Wild One', article by Dick Teresi, Forbes magazine May 1999

If cycle helmets really work, i.e. provide a significant level of protection from brain injury and death from brain trauma, then our debate has to focus on (a) to what extent is cyclist injury a major social cost, and (b) does this justify, on civil liberties grounds, social control on the scale of MHLs. But if cycle helmets do not really work -- or do not work as well as other strategies for improving cycling safety -- or do not work as part of a whole-population public health strategy, then we're arguing about nothing because there is no justification for MHL.

My original long essay which you may have just read (or more likely skipped over) is a discursion of the first kind. It's about risk and society: how do we perceive risk, how do we manage risk, and when does the society, as incarnated in its government, have a right to dictate to individuals how to manage their own risks (i.e. when is individual risk the same thing as social risk). The essay more or less assumes that helmets work, and that risk is significantly or at least measurably reduced by their use. But it questions how major/urgent this risk is as compared to other "public burden" issues; and it questions the propriety of governmental authority's overriding my autonomy in managing my own risks.

In the months since I wrote this essay, I've been reviewing statistics and research that make me doubt the real effectiveness of cycle helmets. See the URLs and soundbites for some of the information which has influenced my own thinking. This places the discussion on new territory. This is the state of play as I understand it today (early 2001).

Most of us road cyclists believe that the primary fatality risk to cyclists, as well as to pedestrians and motorcyclists, is automobile collision. Cyclists may have all kinds of lesser accidents involving trees, potholes, sand, other cyclists, and so forth -- but when we die it is usually as a result of collision with a car.

Many of us believe that the primary risk to ourselves and to pedestrians is easily identified: motorists operating motor vehicles illegally, incompetently, or irresponsibly. On the other hand, some cycle advocates believe that cyclists are more often at fault than drivers, and that better cycling skills would remedy most of our problems. But we concur that the primary "social cost" risk we need to consider is collision with a car, truck, or bus. These are the incidents which most often produce death or disabling injury. This is the arena in which we must measure and evaluate helmet efficacy.

Dr George Shively of the Snell Memorial Foundation (the people who test and rate helmets for both motorbikes and pedal bikes) said, ". . . it is impossible to build a helmet that will offer significant impact protection." I suspect that what he really meant was, if such a helmet were built, it could never be sold. It would be so huge, so massively padded, so heavy, that it would not be wearable.

So how are real, saleable helmets built and tested?

I will here quote extensively from the excellent Ontario Coalition for Better Cycling Helmet FAQ, since I could put it no better myself:

What kind of protection does a helmet provide?

Helmets are tested in the lab for straight line (linear) blows only. Test procedures set by standards bodies like Snell, ANSI, and CPSC require a helmet containing a 5kg (11lbs) rigid headform to be dropped onto a flat anvil from a height of 1.5 to 2.0 metres (5ft to 6ft 8in). If more than 300g's is imparted to the headform the helmet cannot be certified.

The helmet's outer shell is designed to mitigate the effects of an impact by spreading the force over a greater area of the head and by reducing friction in a slide. The helmet's liner, however, is made of stiff foam and requires a certain minimum force before it starts to crush. Until this minimum is reached, the head must absorb the impact. Once the minimum is reached, the helmet's liner absorbs energy until either all the remaining energy is absorbed or the liner has been crushed to its minimum thickness. If the blow is of such severity that the liner gets crushed to its minimum thickness, remaining energy is absorbed by the head and is likely to be enough to be lethal. This means for a fall, whilst helmets are designed to limit the impact (deceleration) to a just sub-lethal level, they won't reduce it much below that. In these cases, "sub-lethal" translates into anything from a very bad concussion to a coma. Any impact of greater severity than received in a typical fall will be lethal.

The medical profession now believes that even lesser accelerations can produce serious injury and that the 300g level is too high. However, it is unlikely that helmet standards will be raised to provide significant protection because the industry doesn't believe that consumers would buy the resulting products.

The trend is in the opposite direction. In Australia, the standard was actually lowered because helmets produced under the old standard did not meet with market acceptance. Manufacturers are presently responding to market demand for helmets which improve air-flow inside the helmet and to fashion by manufacturing helmets with more holes them. While these pass standard tests, they spread impacts over a smaller area of the head, so when an impact occurs it will be more concentrated around the center of the impact.

Do industry tests accurately simulate real life cycling accidents?

Yes and no. While the biomechanics of what happens to a cyclist in an accident is very complex, industry tests are extremely rudimentary. Tests attempt to replicate a fall from a bicycle where a helmeted head directly hits a fixed solid object. Because the use of a rigid headform is prescribed, test procedures are unable to simulate the effect on a human head of bone and soft tissue. In the test, the foam liner starts to crush immediately upon impact, not so in a real life. With a real head, because the head is less rigid than the foam liner, the head absorbs impact before the liner starts to crush. Also, tests exclude any possibility of measuring what might happen inside the skull where brain tissue contorts, rotates and tears as the skull is impacted. The result is helmets being manufactured which optimize padding stiffness even though softer padding could provide better overall protection. Reflecting concern over this and the effectiveness of existing helmets, the Canadian Standards Association has altered its test specifications for children's helmets to ensure they are constructed of softer padding.

Are helmets supposed to provide protection against all impacts?

No. Sharp, high speed objects are likely to penetrate helmets particularly those with many vent openings. Also, helmet tests monitor the effect of linear force but not rotational force. A blow which is not square on centre, i.e. not linear, will rotate the head. Diffuse injuries - the most serious and common type of brain injuries - result from rotational stresses on the brain. Linear force on the other hand, result in focal or localized injuries rather than diffuse injuries.

It has not been ruled out that the added mass, size and surface texture of a helmet may make the rotational effect more severe. A "safety" device which has been shown undeniably to assist in rotation and increase the risk of diffuse brain injury is the headrest on car seats in rear-end crashes. Nothing has been shown one way or another though, for bicycle helmets. When acquiring new helmets, buyers should consider helmets which are spherical in shape as they are more likely to minimize rotational effect than the trendy duck-shape aerodynamic helmets.

How about testing for high speed impacts?

The maximum 2 metre (6'8") drop simulates a 20 km/h (14 mph) impact. Direct impacts over 20km/h can be expected to be lethal.

What does all this mean?

For a start, a 14 mph impact is typical of toppling over (fairly slowly) on your bike at low speed, as in losing your balance on ice or sand. Or, typical of encountering a car moving at 10mph while you are moving at only 5mph. Or vice versa. This kind of collision might well happen in a parking lot, or at a driveway or a 4-way stop where there was a misunderstanding.

However, the average cyclist in reasonably fit condition can ride on level terrain at 15mph and more, and the average car in city traffic is moving at 25 to 30 mph during its brief spurts of activity between stop signs or lights. On boulevards the car will be moving at 45, and on highways at 55 to 80 mph. A cyclist on a smooth downhill grade can easily average 30 and touch 45; a recumbent rider or trike jock can see 60 mph. A car speeding to run a red light may be travelling in excess of 40mph even in a 25 zone. A car in an uncontrolled skid may be slewing sideways or spinning at a high speed. I'm told that the impact of a car moving at 30 mph is equivalent to a shotgun blast.

These are the situations in which cyclists most often are killed by cars. We do not usually die in parking lots, or from toppling over all by ourselves. Some of our more reckless brethren and sistren have been known to wrap themselves around trees or grab a little too much air on extreme downhill singletrack, but they are the exception: the average dead cyclist in America dies by impact with a motor vehicle.

Therefore, the news that helmets are only effective in situations where cyclists usually are not at much risk, and that they are ineffectual in more commonplace and more risky situations, is disheartening. Ironically, one often sees a cyclist slogging up a long grade at about 8 mph, with helmet hung on pack or handlebar; same cyclist will don the helmet before whizzing down the other side. Alas, this cyclist has abandoned the hat while travelling at a speed at which it might actually provide head protection, and has put it on again only when his/her velocity renders it useless.

But, you may say, if helmets are this useless, then bike fatalities cannot be diminishing. Whereas "it's common knowledge" that now that we're enlightened and lots of people wear helmets, biking is far safer than it used to be. Well, "safer" is not the same thing as "reduced fatalities".

Riley Geary, who has done a lot of primary research in the US national FARS database, notes (private correspondence) that

Some of the more clueless helmet promoters point to the nearly 70% decline in juvenile cycling fatalities since 1975 as evidence that helmets are indeed having a dramatic impact on such fatalities, but fail to notice that juvenile pedestrian fatalities have declined by essentially the same amount over the same period (actually, a wee bit more). Since no one has figured out how to get pedestrians (or motorists for that matter) to wear helmets yet, clearly the observed decline in juvenile cycling fatalities has little or nothing to do with increased helmet usage, and everything to do with decreased exposure rates (and changing demographics of course)--fewer and fewer kids being allowed to walk or ride a bike anywhere.

By contrast, adult cycling fatalities have increased substantially and now stand about 65% above where they were at in 1975, while adult motorist fatalities have remained relatively stable (currently about 5% above 1975 levels), and adult pedestrian fatalities have declined somewhat (about 20% under 1975 levels).

When pressed on the issue, some of the more knowledgeable traffic safety types will admit that there is no real evidence bike helmets are having more than a negligible effect on fatalities . . .

[See also Post Hoc Ergo Propter Hoc, a paper in progress about this misinterpretation of FARS juvenile fatality statistics.]

So, though the highway safety establishment, police, and other helmet promoters credit the uptake of helmets with a decline in fatality stats, a careful look at real statistics doesn't bear out this comforting thought.

You can wade through all the references in my URLs page -- please do! -- but the bottom line as I see it is that in the US, in the UK, Australia, Canada, and New Zealand, in the twenty-odd years since helmets have become either trendy or mandated or both, there has been no significant mitigation of fatality rates for road cyclists. There has been a helmet effect -- a marked effect -- but it has been the discouragement of cycling.

Most people immediately turn to motorcycling helmets as the authoritative example. We "know" that motorbike helmets work, therefore cycle helmets have to work.

In fact, the effectiveness of motorcycle helmets is also looking less certain than promoters and enforcers had hoped. Those other bikers, the kind with big engines and bad attitudes, also don't seem to be getting much in the way of increased life expectancy out of their plastic hats. In a comparison of US states with "lid laws" and those without, ABATE (an American motorcycle advocacy membership organisation) found:

If helmets actually provided the protection that some legislators and insurers claim, it follows that fatality and accident rates would be significantly lower in states in which helmet use is mandated. In fact, the opposite is true: the motorcycle fatality rate in helmet-law states is 2.97 deaths per 100 accidents versus 2.79 in helmet-free states. This amounts to 18 additional deaths per 10,000 accidents in mandatory helmet states!

. . . According to Doug Fits, manager of the award-winning California Motorcycle Safety Program (CMSP), a comprehensive study of the effectiveness of rider training completed by the independent firm SYSTAN, Inc. entitled California Motorcyclist Safety Program -- Program Effectiveness: Accident Evaluation has determined conclusively that formal training reduces accident involvement; that accidents among under-18 riders have dropped by 88 percent (according to the California Highway Patrol, young riders are the most likely to be involved in serious crashes), while accidents among riders 18 years of age and older have dropped by 61%.

From the foregoing, it appears that no matter how well-intentioned, the legislature was seriously misguided in its attempts to reduce "public burden" by enacting the mandatory helmet law.

-- Linda Gee, ABATE of California

It is not at all remarkable or peculiar that motorcycle advocates came to the same conclusion that most cycle clubs and advocacy groups have reached: the key to road safety is road skills. Second to sound, practised riding skills, is public awareness and convincing motorists to behave themselves and share the road with two-wheelers.

But that was a motorcycle rights advocate. Police often insist quite the reverse, saying that helmets are a must for personal safety. But not all cops think alike. The Commissioner of the CHP, D. O. Helmick, said in August of 1999: "A comparison in accident statistics was made in order to assess the effectiveness of California's law, 1987 through 1991, were compared to the first five years after the law became effective, 1992 through 1996. The average motorcyclist fatality rate for California for the last five was 2.524. The average for the five years after the law, 1992 through 1996, was 2.398."

Theses figures are fatalities per 100 accidents. So, in this official view, the helmet law changed the fatality per accident rate from 2.524 percent to 2.398 percent, which is not significant.

Motorcycle helmet laws in the 1970's were almost universal. During the 1980s most states relaxed the requirement to cover only a youthful age range -- 17 and below in many states, 18-20 and below in others. California held out stubbornly for all ages, as did about ten other states. Colorado, Illinois, and Iowa have no helmet laws for either motorbikes or bikes. You can find a chart of State helmet laws online.

Riley Geary in the course of his long crawl through FARS has made a painstaking compilation of bike and ped fatality rates in different states of the US from 1993 thorugh 1996. They are rated in terms of fatalities per million population. If bike helmets were markedly effective in reducing cyclist fatality, we would expect Iowa, Illinois, and Colorado (states with no law even for juveniles) to be at the top of the list for bike fatality.

In fact, Florida (has a helmet law for juveniles) ranks first (a dubious distinction) with 8.6 fatalities per million people, or 720 deaths in 3 years. Florida, Delaware, California and Oregon all rank above the national average in cycle fatalities, and all have youth helmet laws. Of the remaining states, fewer than half have bike helmet laws. Of the bottom 6 states (those with very low fatality rates), two have youth helmet laws and four do not. The "safest state in the Union," North Dakota, has no helmet law. It also has some darned cold winters :-) maybe its cyclists spend half the year tucked up under the eiderdown at home.

At any rate, clearly we don't see a strong correlation in fatality rates. If we did, the states with low fatality rates would all have helmet laws and the states with high fatality rates would have no helmet laws. It's a lot more random than that.

Of the thirteen states that were above the national average in cyclist fatalities, 11 are also above average in overall traffic fatalities. Four (California, Florida, Arizona, Louisiana) are also above average in pedestrian fatalities. Perhaps people drive badly in these states?

But it's even a little more complicated than that. California is above average in cycle and pedestrian fatalities, but below average in overall traffic fatalities. This may reflect the fact that California has been developed with an eye to accommodating drivers rather than peds or bikes; in sunny (smoggy) CA, the car is king. The billions of dollars CA pours into highway, freeway, and parking facilities, and our deliberate engineering of cities to be car-friendly rather than human-friendly, may result in lower auto passenger/driver fatalities but higher ped/bike fatalities.

1997 analysis of US National Highway Transportation Safety Administration data (to 1995) uncovered no statistically significant drop in cyclist fatalities in the eight states which have implemented MHL for at least one year. In the affected age group, New Jersey's fatalities dropped from four to one in the year following the law, then rose back to five the next year! California's changed from an average of 34.75 fatalities in the four years before the law to 34.5 in the two years since the law. Significant drops in children cycling to school have been reported after the introduction of the helmet laws. When cyclists quit cycling and turn to sedentary lifestyles, the general health of populations decline. Everyone among those populations pays the resulting costs of increased health care.

-- Ontario Coalition for Better Cycling

A twenty-three year plot of bike and ped fatalities in Canada shows no sudden dip in cyclist fatalities during the years when bike helmets were becoming popular, nor when they were mandated. The cyclist fatality rate shows the same slowly declining trend as the ped fatality rate, probably due to "programs aimed at motorist behavior" as the OCBC suggests. A similar graph for the US shows the same pattern.

In Australia and also in New Zealand, cyclist fatalities have not been measurably reduced by either the trendiness of helmets or their mandated use. What has been reduced in each case is the number of cyclists on the road.

This is echoed in the US by the figures on motorcycle registration. A cycle club newsletter from Texas notes "In Texas, since the repeal of the helmet law, motorcycle registration has increased by 6,900, bringing the total registrations to over 240,000 riders. The Texas Motorcycle Rights Association, realizing the mistake that was made, has become aggressive in their political work and now have members that total 42 percent of the delegates in the State Democratic Party."

Just the raving of some libertarian, leather-clad biker? Actually, a far more formal analysis of CDC motorcycle safety statistics concludes: "The number of motorcycle registrations per 1000 population is dramatically higher in states without a comprehensive law--in these states, people are more likely to own/register/ride a motorcycle. Thus, motorcycle usage per capita is higher as will be fatality rates. In particular, in states with comprehensive laws, there are 18.7 registrations per 1000 population during 1979-86, in states with no laws there are 35.9, in states with partial laws 33.9 and in states without a comprehensive law (states with no laws or partial laws) there are 34.6."

This author goes on to comment "One way to correct for the influence of higher registration rates is to calculate a registration-based fatality rate associated with head injuries. The CDC study does this and finds that the differences between comprehensive law states and other states all but disappear. In particular, the comprehensive states have 3.0 fatalities per 10,000 registrations while partial law states and no law states have 3.7 and 3.5 respectively."

And, more amusingly, "Another disturbing aspect associated with the registration based rates is that the CDC study ignores the fact that partial law states have a higher fatality rate than states with no laws. Using the same flawed logic contained in the CDC study, one could conclude from this result that helmet laws do not work because they result in higher fatality rates. Yet the study just ignores this finding. Of course, the contradictory result supports more the notion that the CDC study has produced distorted estimates of helmet law effectiveness than it supports the notion that helmets cause fatalities."

When statistical analyses are done shoddily (as "safety experts" do when they compare fatality rates per million population rather than fatality rates per thousand registered motorcycles), they ignore possibly relevant information (such as twice as many motorcyclists living and riding in states without helmet laws). The Australian government did likewise when it announced that its mandatory helmet law had reduced cyclist fatalities. It had, but not by a percentage any larger than the decline in the number of cyclists. One way of preventing cycle injuries is to discourage cyclists from riding at all.

There is a school of thought among motorcycle MHL opponents (as quoted above) which bears consideration. These skeptics note that in the US and worldwide it is often individual motorist insurance companies, or cartels of insurance companies, who lobby most strongly for mandatory helmet laws. They also note that in most fatality collisions between motorcyclists and drivers, the driver is at fault, and the driver's insurance company can take a heavy hit in compensatory claims.

An obvious solution from the insurer's point of view is simply to get the motorcyclists off the roads. The most vulnerable road user is the one your client (possibly a clueless and/or aggressive driver) is most likely to do expensive damage to. Therefore, to protect your bottom line, discouraging the use of the roads by more vulnerable transport mechanisms makes sense. MHL's have had a marked discouraging effect on motorcycle registrations in the states where they have been enacted. From the insurance company point of view, that's a win: fewer claims.

We could easily extend this line of reasoning to bicycles and pedestrians. Being unarmoured, and slower than the dominant automobile, bikes and peds are at a disadvantage as long as our road system is run like some kind of experiment in Social Darwinism. Why not reduce the whole liability question by discouraging bikes and peds: mount massive "safety" campaigns actually designed to scare people out of cycling or walking, and lobby for mandatory helmeting of cyclists. MHLs for juveniles have measurably reduced juvenile cycling already in states such as California, so why not keep at it and get more cyclists off the streets so that motorists can drive dangerously in peace?

If we keep on in this vein, eventually we end up banning ragtop sports cars -- and indeed all small, light vehicles -- and promoting the sale and use of ever-huger, ever-heavier "armoured cars". Gee, isn't that what is already happening in the US with the craze for gas-guzzling SUVs? At any rate, this simple-minded equation of size, weight, and armour plate with "public safety" is something to watch out for.

So the discouraging results of actual helmet use in the real world leads us right back into the philosophical realm, and our questions about why governments of wealthy nations like the US, Canada, Oz, and NZ are so willing to force cyclists to use "safety gear" which can pretty easily be shown to be ineffectual. Some say it is a massive conspiracy between the plastics industry, the helmet manufacturers, and the governments. I prefer to finger the insurance companies, but there's no question that laws which mandate purchase of "safety" equipment do provide manufacturers with a great free marketing program. I'm sure there is a little enlightened self-interest operating all round!

But there's a fundamental reason why the helmet ideology is so acceptable, so easy to sell despite shoddy science and contradictory real-world results. As I argue in my essay, I think it is far more a cultural obsession, a belief system based in the centrality of the automobile and the devaluation and marginalisation of all other forms of transport.

Many cyclists who do wear helmets are convinced that their personal safety is enhanced by so doing. My feeling is that they should have that choice. However, the data we have after 25 years of helmet promotion and take-up are in my opinion at best inconclusive, not a substantial justification for government promotion or mandate. If we are less charitable, we can find evidence aplenty to suggest that by harping endlessly on the subject of helmets and danger, the road safety establishment managed to "dangerize" cycling so successfully that millions of people simply stopped doing it, stopped allowing their children to do it, and are now driving their cars everywhere . . . thus making the roads more hazardous for the few stubborn pedestrians and cyclists who continue to use public space.

Cycling is perceived as dangerous partly because we cannot bear to face how dangerous our automobile culture is. US drivers are killing each other, as one cyclist said, at a rate that's "the equivalent of a jumbo jet crash every 2 or 3 days". We would rather natter endlessly about how dangerous cycling is, and force cyclists to wear ineffectual safety gear, than face the cost of our automophilia. Addressing the pressing social, economic, and environmental problem of automobile saturation and overuse, though far more challenging than simply telling people what to wear, would do a lot more for the safety of all road users: cyclists, pedestrians, drivers . . . who knows, we might someday see kids playing kickball in the streets again. And that would be a happier world.

The Australian enforcement doesn't seem a very encouraging example, even if one accepts the milder version of ridership decline statistics. Australian government representatives and apologists for the helmet law proudly point to decreasing bike fatality figures over the years since the enforcement; but Dorothy Robinson's research paper (see URLs) shows a very small, or even negative (weighted by decreased ridership) impact on bike head injury fatalities. She contrasts this with a very strong correlation between a general reduction in pedestrian, bike, and auto injury/fatalities and a shift to stronger DUI enforcement and traffic control.

We should perhaps note that the Australian government, having apparently discouraged large numbers of riders, then spent over $30M on bike trails and bikeways to try re-encourage them. Whether this is an admission of guilt or just a progressive trend is another vexed question. Just to muddy the water, there's ample evidence accruing that cycle paths may not be the key to safety either!

Governments of affluent industrial states keep looking everywhere for the secret to road safety . . . helmets, cycle paths, school education programs, "management" of pedestrians and cyclists. Like the Purloined Letter, the real cause of road danger is in plain sight, yet they cannot bear to look at it: overuse of the automobile.