Are helmet laws suppressing cycling?

If maps were based on time not distance, this is how big (and small) the Netherlands would look

A new Australian study has thrown more fuel on the fiery debate about whether or not bicycle helmets should continue to be mandatory. Its headline claim is 23% of Sydneysiders say they would cycle more if they weren’t obliged by law to wear a helmet.

This isn’t merely saying that some people would prefer to cycle without a helmet – it’s claiming the law actually suppresses cycling.

I lean toward the school of thought that says mandatory helmets probably do more harm socially than good, but as I’ve said before, it’s not the sort of issue that I would want to die in a ditch over. However if there were reliable evidence that compulsory helmets actually restrain cycling, that would require a rethink.

The research was undertaken by Professor Chris Rissell and his colleague, Li Ming Wen. It is published in the latest edition of the Health Promotion Journal of Australia, under the snappy title, The possible effect on frequency of cycling if mandatory bicycle helmet legislation was repealed in Sydney, Australia: a cross sectional survey.

It’s a brief and easy to read article but a summary by Professor Rissell was published on The Conversation last week, Make helmets optional to double the number of cyclists in Australia. Professor Rissell is a self-confessed cycling advocate and firmly in the activist “repeal” camp on helmets.

He and his colleague surveyed 600 Sydney residents aged 16 years and over. They found one fifth of respondents “said they would cycle more if they did not have to wear a helmet, particularly occasional cyclists”. They conclude that:

While a hypothetical situation, if only half of the 22.6% of respondents who said they would cycle more if they did not have to wear a helmet did ride more, Sydney targets for increasing cycling would be achieved by repealing mandatory bicycle helmet legislation. A significant proportion of the population would continue to wear helmets even if they were not required to do so.

Regrettably, I don’t think this study adds anything to our knowledge of whether Sydneysiders would ride more if helmets weren’t compulsory. They might, but then again they might not. The trouble is the survey relies on a hypothetical situation: “Would you cycle more often if you didn’t have to wear a helmet? Yes or No?”.

Hypothetical survey questions are notoriously unreliable. I’m not picking on the anti-mandatory helmet brigade here – I also took Metlink to task earlier this year for trying to make grandiose predictions about future public transport patronage based on a similarly unreliable methodology.

It’s standard practice to avoid hypothetical questions in surveys. Consider this advice from The World Bank publication, The Power of Survey Design:

Hypothetical questions, especially, should be avoided. People cannot reliably forecast their future behaviour in a hypothetical scenario. Thus, the questionnaire design should make careful use of this style of question.

Hypothetical questions are especially problematic when respondents are asked to predict an activity they’ve had little experience of. The Canada Business Network advises questionnaire designers, if possible, to “avoid hypothetical or future intentions questions:

Hypothetical questions force the respondent to provide an answer to something he or she may never have thought about and, therefore, the respondent may not be able to provide an accurate response.

The authors should’ve been alerted that all might not be right when they found 40% of those who say they’d cycle more if helmets weren’t compulsory, also say they support mandatory helmet legislation. Yes, there’re scenarios where it’s conceivable someone could hold both views simultaneously, but 40%?!

The authors implicitly concede hypothetical questions are of little value when they say “the relationship between intent to perform a behaviour and actual behaviour (for example, helmet wearing) is not known”. And they effectively throw in the towel when they say “if only a half or even a quarter of these people who say they would ride more actually did so, our data suggest that the Sydney cycling participation targets would be easily reached”. A “quarter” is just 6% of the sample! In fact there’s no rhyme or reason for choosing “if only half or even a quarter” in preference to, say, “if only a tenth or even 1%”.

The survey was part of an omnibus telephone survey covering a range of topics, so the researchers only got to ask four questions specifically about cycling. This limited their ability to investigate what respondents really meant by their answers to the hypothetical question. There are also some technical issues with the study like leading questions and no data on the response rate.

So, closer inspection shows there’s nothing in this study that persuades me mandatory helmets are a key obstacle to wider cycling in this country. They’re not helpful, but they’re not the main problem either. There are some other interesting aspects reported by the study which I hope to come back to another time.


Book giveaway: follow this link to be in the running for one of two copies of Jarrett’s Walker’s new book, Human Transit

59 Comments on “Are helmet laws suppressing cycling?”

  1. Dave says:

    I’m not sure that 40% is unreasonable to hold both views. I would cycle more if helmets weren’t compulsory, but fully support the mandatory nature – I think individuals often underestimate the risk of serious injury, and further don’t price in the costs to society should they incur a serious head injury (or death) as a result of having an accident without head protection.

    Similarly, I’d drive a car at 200km/h to get to somewhere quicker if that were allowed, but it doesn’t mean I don’t support having speed limits more appropriate for the circumstances, because my assessment that 200km/h was safe might not be that accurate… ; )

    • Alan Davies says:

      I think you’re probably exceptional, Dave 🙂

      • Michael says:

        It might be interesting to test questions based on who the respondent had in mind when answering the question. People might honestly answer that they would cycle more if helmets weren’t compulsory. I have on the odd occasion thought that if the bike share scheme didn’t require a helmet I might give it a try.
        I do however wear a helmet when I ride to work and I feel safer wearing one. This view is not based on a full analysis of the data or an objective assessment of risk, it’s just habit. If I imagine kids riding I would probably answer that helmets should be compulsory even though I and my friends all survived childhood without helmets. I suspect the original motivation for compulsory helmets were to marginalise cycling and make cyclists responsible for the danger cars poised to cyclists. Few people are really equipped to give a thoughtful answer regarding compulsory helmets.

  2. Jim Wright says:

    I suspect that the dilemma of wearing or not wearing helmets is related to the cycling styles in Australia and say, the Netherlands, which is often quoted as the model cycling community. In Australia, most cycle lanes run alongside of roads carrying motorised vehicles of one sort or another. Unless cyclists and motorists both behave with the greatest care, there is a non-zero chance that cyclists will be injured on account of the traffic conditions. This is not helped by the fact that the Australian cycling image is that of the competitive cyclist, who thinks that he or she (but usually he) thinks they are Tour de France material and ride accordingly. Not for nothing are the extremists referred to derisively as the Lycra Lunatics. In the Netherlands, on the other hand (and I have seen it myself), cycling is a much more sedate affair and more often than not, confined to dedicated tracks. Hence, when trying to decide whether or not we should wear helments, we are comparing two entirely different situations. The answer seems to be – wear a helmet if you want to travel along heavily trafficked roads at high speed, but feel free to leave your helmet at home if you wish to travel sedately along a path in a park.

    • All that you’ve said is true, but you’ve also described a chicken and egg scenario.

      When helmet laws were first introduced in Australia there was an immediate and considerable drop in the numbers of people that cycle. This fact isn’t disputed by either side of the debate. I would hypothesise that the “casual rider” formed the bulk of this group. I would also argue that as a result of the lower cycling numbers, there has been less investment in cycling infrastructure which has further suppressed demand for the all cyclists, but especially the casual commuter.

      I think Darwin has taken the best step that could be taken in other Australian cities. Repeal helmet laws for all cycling tracks and footpaths and then build a lot more cycle paths!

      If we’re not going to remove the helmet laws, it would at least make sense to legalise riding on the footpath. Something I do regularly if I’m weary of the roads (i.e. Hoddle St, Alexander Pde) in my years of doing this I’ve never once had a problem navigating past pedestrians and I feel far safer for it. Having laws that make it mandatory to ride in the more dangerous space is ludicrous.

  3. kymbos says:

    If you already have a helmet, why would you ride ‘more’ if you didn’t have to wear it? Unlike Dave, I can’t fathom a situation where I’m keen enough to ride my bike, but consider the need to wear a helmet as enough to stop me. It’s ridiculous. People for whom that is true deserve to be stuck in commuter traffic.

    I agree the study adds little to our knowledge. Until we have a natural experiment (like one state changes laws), we’re not going to know the marginal difference the helmet laws make.

    • Dave says:

      It’s mainly in relation to helmet availability. If you’ve always a helmet with your bike, no problem. I work in the city and am surrounded by the hire bikes, and I would use them a lot more for spontaneous trips if I didn’t have to source a helmet. I can make trips from the office (helment there) but not from other trip origins, unless I buy a 7/11 hardtop.

    • First and foremost there is the obvious bike-share situation, but there are other situations where access to a helmet can be limited.

      Borrowing someone else’s bike where the helmet doesn’t fit; or if you are a guest at someone’s house where there is a spare bike, but no fitting helmet. This happens regularly at my house where guests come from out of town often and despite that we have the same number of helmets as bikes, we don’t carry every helmet size!

      Finally there are other times where donning a helmet just seems unnecessary. If I ride the 2 minutes up the road to the shops is it really that important that I put the helmet on?

    • Katie says:

      Hi Kymbos,

      I’m a very keen cyclist. I cycle to and from work every day. I’ve enjoyed cycling for most of my adult life.

      I would dearly love to cycle more, and abandon car travel altogether. But the helmet is very often the very thing that stops me from making ad-hoc trips on the bike.

      For example, tonight we are going out to dinner at a nice restaurant, 5km from home. Perfect for cycling. But we will be driving. Driving will take longer, cost more, and parking will be a nightmare.

      The reason: I don’t want to mess up my freshly washed and styled hair, by donning a helmet. I get very sweaty under my helmet, especially in these warmer months. If I didn’t have to wear the helmet, it would be a no-brainer to cycle tonight.

      You might say it’s “ridiculous” to let such a frivolous thing stop me from cycling. But I’m a woman, and have long hair, I enjoy going out to dinner with friends, and like to look nice when I go out. Sometimes that wins out over being some sort of cycling hero.


  4. Chris Rissell is an anti-helmet (law) activist. He refuses to wear one. He very actively advocates the overturning of the mandatory helmet laws. I think he does himself no favours by also being involved in related academic studies … I know his biases and don’t believe they don’t colour his research.

    To me it’s a question of political priorities and where bicycle advocacy time is best spent. I honestly cannot see these laws overturned … if they were sooner or later some uni student will be in a crash and end up with devastating brain damage. And in a second the media, the family, etc will be looking at the politicians saying why, oh why, did you change those laws? No pollie wants to be on the receiving end of those questions.

    Cycing is growing steadily and exponentially in Sydney as it is — lets find realistic ways to keep that going.

    • Every academic worth their salt is an activist in their field, that’s part and parcel of the job description. What would be the point in spending your life studying if you weren’t pushing for what you believed in? Furthermore if you were an activist with an academic background why wouldn’t you use your academic talents to research your interests?

      I can’t see the laws overturned in one movement. I can see other cities following the NT example. Especially as it currently stands per capita cyclist injury statistics are on par with the other states, despite the higher per capita motor vehicle accidents in the state.

      • Alan Davies says:

        “Every academic worth their salt is an activist in their field, that’s part and parcel of the job description”

        Actually no, I think the idea of the “ivory tower” is the opposite. If academics don’t strive to be disinterested then there’s the risk their political views will colour their judgement. For activists and advocates, persuasion becomes the objective, not knowledge. And truth is the first casualty of persuasion (military or otherwise).

        That anyway is the ideal.

        Addendum: meant to add that I think the helmet laws will inevitably be repealed when people think it’s safe to cycle – not the other way around.

        • Again, I’m going to disagree strongly with you. One can both pursue knowledge and express opinion based on the knowledge they have gained. You do it on this blog, as does every commentator. To claim that any of us are “disinterested” is absurd. I’m not trying to claim that Chris Rissell is a “good” academic, nor that his research isn’t flawed. The extent I’ve read of it is the introduction and discussion of the paper you’ve linked to above. However if academics do are not “active” in their field what purpose does their research serve?

          Lectures, academic journals, interviews with the media, consulting interested parties. All of these are part of the academic life and all are forms of activism. None are disinterested, but that does not mean that outside knowledge and different opinion should not be pursued.

          • Alan Davies says:

            My first point is that the traditional conception of a university is of an “ivory tower”. Whether one likes it or not, that’s a matter of history and observation.

            The second point is whether or not that’s sensible or even realistic. It’s true there’s no such thing as perfect objectivity, but it’s possible to strive to be disinterested. It’s also possible to reconcile a concern for the real world application of a discipline with a strong degree of “disinterestedness”.

            If I had a bad disease I’d want medical researchers to be concerned and motivated to find a cure. But I’d want them to strive to be disinterested – I wouldn’t want them to be biased by personal glory or by some drug company’s blandishments. Nor would I want them to be so convinced of the rightness of a particular treatment that their minds were closed to new evidence (I recommend reading the Emperor of All Maladies on this score).

            Strong advocates run the risk of being close-minded because their mission is to persuade. Helmet law is a more overtly political issue than research on a specific disease, so it’s more vulnerable to abuse.

    • Prof Rissel’s and colleagues researched the effect of helmet laws and found that they didn’t reduce head injuries. Other researchers also examined the head injury data, coming to similar conclusions –

      If research shows that a law is bad – doesn’t reduce head injuries, but instead adds to healthcare costs by reducing cycling by 30-40% (leading to reduced safety in numbers) – isn’t the only decent, honest option to oppose the legislation that has been shown to be detrimental?

      Cyclists are still suffering devastating brain damage. If you do the stats, the risk seems to be higher than what would have been expected without the helmet law.

      And, while the downward trend that started with helmet laws is finally beginning to turn around, I’d hardly say there’s been exponential growth – see the graph of census data on cycling to work

      Given the big drop in cycling with helmet laws, and the tiny increase here compared to much greater increases in cities such as London and Paris (where they don’t have helmet laws).

      I spend a lot of my bicycle advocacy time on local issues, because I want the roads where I ride to be safe for cycling. But there’s no doubt that the biggest single issue that reduces my safety is helmet laws. If we could return to pre-law cycling levels, there’d be much greater safety in numbers, and reduced risk of serious brain injury from collisions with motor vehicles, and less need for the “Sorry I didn’t see you mate” excuse.

      • Dave says:

        The census takes place in the middle of winter, which is probably capturing cycling at one of its lowest points for the year. Further, the census asks what the primary mode you took to work was – if you rode 5km to the station, then got a train 40km, you’d have to answer the train as your primary mode.

        However, it’s still useful data, especially when combined with the spring ‘cordon counts’ often conducted to measure cyclist numbers.

        • Actually, if you ride 5 km to the station and travel 40 km by train, both are recorded – the trip is counted as a multi-mode journey. In 2006, 4.5% of work trips were multi-mode, but there were so few train/bike and bus/bike trips that they fall into the ‘other’ category in the standard files.

          So the percentage of single mode journeys by bike as percentage of total single mode journeys should give a reasonable n idea of the changes in cycling. Being in mid-winter, it’s probably capturing the keen cyclists who are less likely to be deterred by helmet laws than casual cyclists.

          There was a very interesting analysis of other cycling trips – for shopping and education in WA in “Empty Cells, Damned Half-Truths and Pseudo-Statistics:The Lot(tery) of the Bicycle Planner” by Ian Ker. Between 1986 and 2006, there was a small fall in cycle trips to work, from about 1.5 per weekday per 100 people to about 1.0. However, education trips fell from about 8.2 per weekday per 100 to about 1.7 and shopping trips from about 5.2 per weekday per 100 people to about 2.3.

          Comments such as: “If you already have a helmet, why would you ride ‘more’ if you didn’t have to wear it? … I can’t fathom a situation where I’m keen enough to ride my bike, but consider the need to wear a helmet as enough to stop me. It’s ridiculous. People for whom that is true deserve to be stuck in commuter traffic.” suggest to me that current cyclists (who obviously enjoy cycling and don’t mind wearing helmets) forget that helmet laws discourage casual cyclists and potential future cyclists. Helmets are a lot of bother if you have to carry them around when you get to your destination.

          Helmet laws might not be the main reason why people don’t cycle, but I can’t see governments spending oodles of money building new cycling facilities. A few cycle lanes or cycle paths won’t make a great deal of difference to the amount of cycling. In some cases, it concerns me that the things they build are so badly designed that I’d rather ride on the road.

          Realistic options for encouraging cycling are therefore: repealing helmet laws, making vehicle drivers liable for injuries unless there is evidence that the cyclist was at fault, creating more 30 km/hr zones and perhaps allowing footpath cycling if the adjacent road is more dangerous (e.g. has squeeze points or a speed limit of more than 50 km/hr) with cyclists considered at fault in the case of collisions with pedestrians.

          It’s just that people say they would ride more if helmet laws were repealed, but the big drops in cycling when helmet laws were introduced, the reduction in safety in numbers leading to increased fears about safety – all suggest that we won’t see anything like the resurgence in cycling that is starting to happen in other countries unless helmet laws are repealed.

  5. Jim Wright says:

    I run a blog on sustainable population and climate change and as part of my research, I have been trawling through the hundreds of responses to a government issues paper on Sustainable Population Strategy. Among them were papers from two departments in the same university drawing very different conclusions on population demographics. What conclusion should I draw from that ?

  6. David Allen says:

    I’d be satisfied if they made helmets optional for bike hire schemes like Melbourne’s so we could see if it was the helmets that were holding the scheme back. My tip is that the scheme would be extremely popular.

  7. […] and development issues with a particular focus on Melbourne, Australia Are helmet laws suppressing cycling? […]

  8. I am a cyclist since I am three years old…. my mother just pushed me on a small two wheel bicycle along this little lane with gravel and it were my knees first of all that did get hurt in the subsequent learning. I am living mostly in Amsterdam which is a bicycle city optima forma …. cyclist wearing head gear are a seldom sight here (mostly American tourists that took their own equipment).
    Of course it is first of all the non-cyclist motorised participants in city traffic that bring the idea of need for wearing protective head gear while cycling. The imbalance of cyclists is hardly ever a cause of accidents even with an increasing stream of cycling tourists on rental bikes that are pretty bad in doing this propulsion and balancing act.
    We once coined this expression in Dutch “een fiets is iets maar bijna niets” (a bicycle is something but almost nothing) and that saying says it all, keep it minimal. Sadly industry and fashion try to sell products and push them upon us in any way imaginable…. hence the idea of the head gear for cyclists.
    I have cycled in many parts of the world and I would say for instance that I would like to have a harness while cycling in London, would discourage anybody to cycle on country roads in Poland (truck drivers think it is fun to drive you off the road), last cycling in Tokyo is like a ballet with courteous mixing on the pedestrian side walk of those on foot and on two wheels.
    A friend of mine in Paramaribo (Surinam/South America) recently proposed forbidding ‘air bags’ in cars which would make car drivers less adventurous, instead he proposed a pointed steel needle fixed in the middle of a steering wheel, as the best safeguard for Paramaribo cyclists.

  9. John Burke says:

    The interesting thing is that this study is a survey. People are asked whether they would be more inclined to ride a bike if they didn’t have to wear a helmet. 23% of those asked said they would but in other studies on the issue, bike helmet laws are estimated to reduce bike usage by 30%-50%. I’ve never seen an estimate as low as 23% ever and I’ve seen a few surveys on this topic so it is interesting to compare respondent’s expectations with reality.

    Still cycle-helmet legislation and discussion is like religion. It is a fact free zone in general especially any studies supporting helmet use. They are almost non-existant whilst those against are counter-intuitive and hence difficult to understand for many people, exactly like a religion. If you believe you will be paying hospital fees for someone because they didn’t wear a helmet, it will be hard to dissuade you of that fallacy is my experience.

    The real economic win though is in spending on cycling infrastructure, getting rid of helmets and saving money currently wasted on roads.

    • Alan Davies says:

      I don’t think there’s consensus on whether or not Australia’s mandatory helmet laws reduced cycling in all age groups and whether any change observed was permanent. For example, this is from a paper by Walker et al published earlier this year in the journal Accident Analysis and Prevention:

      While the reduction in numbers of teenaged cyclists has been widely cited in the few years immediately after legislation, the opposite was observed among adults and the estimated overall change in cyclist numbers in NSW was close to zero (Walker, 1990, 1991, 1992; Cameron et al., 1992, 1994; Smith and Milthorpe, 1993). It is unknown whether the reductions or increases were temporary or a permanent phenomenon; however, assessments of the legislation’s efficacy must take such fluctuations into consideration. A decline in cyclist numbers was only noted in one of the North American studies (Carpenter and Stehr, 2010).

      Then there’s this study by Haworth et al published last year by the Centre for Accident Research & Road Safety Qld:

      A large study in Melbourne collected data from a number of sites prior and subsequent to the introduction of the helmet wearing laws. This research demonstrated a doubling in the use of bicycles by adults in metropolitan Melbourne. However, there was a decrease in the use of bicycles by children. A decrease in cycling exposure of 10% was observed in children (5-11 years) and an even larger decrease of 44% for teenagers (12-17) (Finch et al, 1993).

      It’s very important I think to look at the bona fides of the various contenders in the mandatory helmet debate. The arguments of zealots/evangalists (for either side of the argument) should be approached with due caution.

      Here’s a link to the 1993 study on Victoria that’s widely quoted – Fincher et al, Monash University Accident Research Unit. There’s also a study by Smith and Milthorpe published by the NSW Roads and Traffic Authority in 1993 that I can’t find anywhere. Pity, because it’s heavily relied upon in these debates.

      • John Burke says:

        Alan you need us zealots and evangalists so as to contrast your own more thoughtful and evenhanded analysis. However at the same time not every issue of contention can be solved by assessment of various studies. For example in the The Great Helmet Debate, we also need to look at studies comparing the rapid uptake in cycling in Europe at roughly the sametime it declined in Australia. I don’t know of any.

        Also since nearly all Europeans think mandatory helmet laws are absurd and laughable and Australian pro helmet supporters no doubt consider Europeans wild risk takers, or stupid in which case an ordinary IQ test between the 2 continents to compare the results might be another satisfactory way to settle the affair.

        I’m an old hand at zealotry and though I’ve always had disdain for compulsary helmet wearing, I’m actually quite new to some of the more sophisticated arguments against these laws and thus claiming my advocacy is based on rational reflection rather than tribalism or something of that nature.

        • Alan Davies says:

          Indeed, there’s a lot that hasn’t been studied and some of what has is of questionable quality. So ignoring studies and relying solely on logic, no one has yet even attempted to dispute the argument I laid out here i.e.

          The first thing the repeal advocates should ask themselves is this: why are only 1% of trips in the UK taken by bicycle even though helmets aren’t mandatory in that country? That’s no better than here! Or why is cycling’s mode share only slightly better in Ireland and Canada than it is in Australia, even though these two countries don’t have mandatory helmet laws? Clearly, whatever the explanation is for the comparatively low rates of cycling in these countries, it has nothing to do with any compulsion to wear a helmet.

          They should also ask themselves why there are such enormous differences between countries where helmets aren’t mandatory. The fact that bicycle use is more than twice as high in the Netherlands as it is in Germany – and nine times higher than it is in France and Italy – suggests pretty clearly that there are other highly influential factors affecting the propensity to cycle that have absolutely nothing to do with helmets.

          Helmet policy doesn’t explain why bicycles capture 34% of trips in Munster, but 13% in Munich. Or why the corresponding figure for Groningen is 37% compared to 10% in Heerlen; or 20% in Bruges but 5% in Brussells; or 19% in Salzburg but 3% in Wien.

          • I said it last time on the issue too, but given the similar rates of cycling between Australia and the UK I’d also like to see what sort of cycling rates have been recorded in London after the introduction of their bike share program alongside the equivalent stats in Melbourne.

          • John Burke says:

            Of course helmet policy isn’t everything in cycling uptake and infrastructure might explain individual levels between cities.
            But as for Ireland, cycling has only been actively promoted since the initial GFC. (An interesting point in itself) Your figures are from 2008 and anecdotal evidence suggests a huge lift in cycling in the past few years, with government subsidies for purchase.
            Pretty much the same in England, where car usage has dropped dramatically since 2008, yearly and exponentially as reported only this week, especially amongst the younger demographic. Cycling has been growing quickly in Britain during the same period.
            As for Wien, you can’t expect big numbers from a town with a name like that.

      • Figure 2 near the beginning of the report you refer to above (Finch et al.) is misleading. It omits the pre-law count for adults in May 1990. You can find this in the peer-reviewed paper at

        When you add in the missing data point for adult cyclists counted pre-law, the pattern for adults is similar to teenagers. There was a substantial increase – from 1069 adults counted in Dec/Jan 87/88 to 1567 in May 1990 at the same sites and observation times, then a fall to 1106 post-law in May 1991. Note also that the value for May 1992 was inflated by a bicycle rally passing through one of the sites – 451 cyclists were counted at that site in 1992, compared to 72 in 1991 – so the apparent increase in 1992 may not be genuine.

        The fact that information on the number of adults counted in May 1990 is hard to find suggests that research organizations were under great pressure to “massage” the truth. Think about it. Comparing estimates of cycle use at different times of year (Dec/Jan 87/88 vs May 1991) is dicey at the best of times. Why would you do this when you have comparable data for numbers of adults counted in May 1990 and 1991?

        In reality, 29% fewer adult cyclists were counted at the same 64 sites and observation periods in May 1991 than pre-law May 1990. It’s disappointing that the Haworth report claims: “In Melbourne adult cyclist numbers doubled after the helmet legislation was introduced but there were fewer child cyclists, particularly teenagers.”

        An insight into the truth can also be seen by looking at the hospital admission stats for head and non-head injuries – at There’s an obvious reduction in *non-head* injuries coinciding exactly with the helmet law. Obviously, if we had no cyclists, there’d be no cycling injuries. The fall in non-head injuries is exactly what you would expect if the law had discouraged cycling. In fact, MUARC analysed the percentage of cyclist admissions that had head injuries, concluding that it was no different to what would have been expected from pre-law trends! (Newstead et al., MUARC Rpt 75)!

        You don’t have to be Einstein to work out if the percentage of cyclists with head injuries was not significantly different from pre-law trends, but there were significant drops in both head and (though it was not reported, non-head) injuries with the law, the most likely cause is reduced cycling.

        I encourage you to look at the data, and also census data on cycling to work to help you decide whether helmet laws discouraged cycling.

        Census data on cycling to work show the same pattern as teenage cycling in Melbourne – increases until the law and then a substantial fall, the same pattern that was evident in numbers of adults cyclists counted – 1069 in Dec/Jan 88/89, 1567 in May 1990, then a fall to 1106 in May 1991.

        MUARC Report 76 described the effect of the law as a 40% reduction in cycling head injuries. This has often been taken out of context, leading people to believe that increased helmet wearing was the main cause, despite the fact that the concurrent decrease in non-head injuries suggests that most of the effect was in fact due to reduced cycling.

        For more details, including graphs of head injuries for individual states as well as changes in fatality and head injury rates for other road users see This paper also reports the survey data for child cyclists in NSW from Smith and Millthorpe and Michael Walker. Most people know that cycling varies considerably over the course of a year, due to different weather conditions. The estimated changes in children’s cycling are valid because they were all conducted in April. The so-called change is adult cycling was obtained by comparing October (when fewer people cycle) with April. Sadly, the authors of the recent report in AAP fail to mention this important fact.

        I had hoped, until I read the recent Haworth and Walker reports, that the pressure to “massage” the truth had diminished. It’s of great concern if governments consider spin more important than the health and safety benefits the might come from increased cycling participation.

        How can you believe anything you see in any government-commissioned report, when, despite 29% fewer adult cyclists counted in the post-law survey than an identical pre-law survey, Haworth states: “In Melbourne adult cyclist numbers doubled after the helmet legislation”.

        • Alan Davies says:

          Hang on, the Robinson paper shows the adult count for cyclists in Vic bounced back by 1992 to be 34% higher than it was in 1991 (the figure missing from Finch) and 95% of what it was in 1990 before the law was introduced.

          • But how much of that was due to the rally at one site, where 451 cyclists were counted at in 1992, compared to 72 in 1991? If the rally involved mainly adult cyclists, it would explain the entire “bounce”.

            There’s also the large reduction in non-head injuries to explain, coinciding exactly with the law. What else could it have been caused by except reduced cycling?

            Finally remember that Haworth claimed that adult cycling doubled. Are you suggesting that it’s OK to describe a 5% reduction in adult cycling as a ‘doubling”?

          • I forgot to add that kids are future adult cyclists. Despite the the fact that a bicycle rally passed through one site, 45% fewer teenagers were riding two years after the law. Surely people who don’t ride as teenagers are less likely to ride as adults?

          • Alan Davies says:

            Robinson shouldn’t have tried to cherry pick like that. The original work was (hopefully) based on a random sample. Perhaps there was a rally back in 1990 too, or maybe 1991 was bad weather, etc, etc. That’s why these sorts of studies require lots of observation points. The author can’t suddenly decide to single out an inconvenient fact in someone else’s research. Anyway, maybe the only reason there was a rally was a direct result of the change in helmet law. Who knows?

            The large reduction in head injuries could’ve been explained solely by the large reduction in teenage riders.

            BTW my mention of Haworth et al wasn’t to either praise or demean their work. It was to show that there are different points of view.

          • Robinson reported all available counts for 1990 (pre-law), then 1991 and 1992. Why do you think that’s “cherry picking”?

          • Alan Davies says:

            The reference Robinson and you both make to the rally is the “cherry picking”. You’re trying to say the 1992 adult data, which shows a big rebound in cycling two years after the new law was introduced, was due to some atypical event and therefore can’t be relied on to show riding didn’t decline significantly among adults.

          • Sorry, I’ll rephrase that. Robinson reported all available counts for 1990 (pre-law), then 1991 and 1992, including all the cyclists counted in the rally. Why do you think that’s “cherry picking”?

            You still haven’t answered why a 5% reduction (including the rally data) in 1992, or something like a 29% reduction in 1992 (rough estimate excluding the rally data) can be described as a doubling of adult cycle use.

    • The 23% is of the entire population, not just cyclists. Only 17.8% actually rode a bike in the past month, so if you add the 23 out of 100 people who would ride more to the 17.8 who did, you get an increase of 129%.

      Of the people who rode a bike in the past month, 37% said they would ride more without helmet laws. Of the 82.2% who did not ride in the past month, 20% said they would ride more. This number of people equals 90% of the number who actually rode a bike in the past month.

      The Cycling Promotion Fund survey noted that 515 respondents were not interested in cycling for transport. Of these, 81 cited as a reason that they did not like wearing helmets. Only 158 cyclists cycled for transport in the past month – 81 more transport cyclists would represent a 51% increase, on top of the 16% of current transport cyclists who said that they would cycle more if not required to wear a helmet – a potential 60% increase in cycling for transport.

      The sad fact is that many people who don’t like helmets appear to have given up cycling, or never took it up in the first place.

      Current cyclists are in general those that don’t mind helmets, at least not enough to give up cycling. They seem to have a hard time understanding why other people – the sort of people who cycle for transport in other countries, including users City Bike Schemes – don’t ride here. This leads to a failure to appreciate the benefits that safety in numbers could bring.

      Seriously debilitating head injuries result mainly from collisions with motor vehicles; many involve forces greater than the design limit for bicycle helmets. I get the impression that regular helmet wearers over-estimate the benefits of helmets and don’t understand that the increased risk of head injury from reduced safety in numbers. Comparison of head injury rates per cyclist before and after the helmet law imply that the benefits of safety in numbers is actually greater than the benefits of wearing a helmet.

      If most cyclists understood this, surely they would support a repeal of helmet laws. After all, you can still choose to wear a helmet, but you can’t get the additional benefit of safety in numbers without repealing the laws, or spending huge amounts of money trying to encourage cycling without repealing them – and with current spending cuts I can’t see that happening.

      • Alan Davies says:

        I don’t think you can add the 23% who say they would cycle more, to the 18% who actually did cycle in the past month. That’s double-counting. Some of those in the first category were also in the second category – in fact the researchers say people who currently cycle are more likely to say they would cycle more if they didn’t have to wear a helmet.

        However, the key issue is that the safety in numbers argument depends on the premise that mandatory helmets really do suppress cycling to a significant extent (as distinct from, say, traffic conditions). I must say I’m beginning to have doubts about that proposition.

        I think I’d still oppose mandatory helmets for other reasons, but I still see it as a marginal issue.

        • I have a copy of the paper and was interested enough to split the results into
          a) a people who cycled in the past month (17.8%)
          b) people who did not (82.2%)

          37% of group a) said they would cycle more if they didn’t have to wear a helmet
          16% of group b) said they would ride more.

          To illustrate this, suppose we are dealing with 100 people.
          The number in group a) who would ride more is: 37% of 17.8 people = 6.58 people.
          The number in group b) who would ride more is: 16% of 82.2 people = 16.05 people.
          The total number who would ride more is therefore 6.58 + 16.05 = 22.64 people, or 127% of the number who rode in the past month.

          Doing the calculations this way leads to a slightly different answer – a 127% increase instead of 129% – but it clearly shows there’s no double counting.

          I think they are different categories – people who cycled in the past month might go from once or twice a month to more regular cycling.

          People who didn’t cycle in the past month might go from nothing to occasional or lots of cycling, as has happened in other countries with the success of the City Cycle Schemes. Velib was said to have driven Paris ‘cycling mad’. I think it also encouraged people to ride their own bikes as well as use the hire schemes. My impression of Paris traffic was that it was even worse than Sydney, yet their public bikes were used 26 million times from Jan to Oct this year – 87,069 rides per day.

          Or think of the Dublin scheme – 550 bikes used 1.29 million times from Jan to Oct this year – 4,251 rides per day, an average of 7.73 hirings per bike per day Over the same period, Brisbane had 74,534 hirings, so Dublin had17.3 times as many bike trips per day, despite the fact that Brisbane has 2,000 city bikes, nearly 4 times as many.

          In Brisbane, 7,951 people cycled to work on census day 2006 (10,886 in Sydney).

          Helmet laws appeared to have an even bigger effect on numbers of cyclists (at least for census data on cycling to work) in non-capital cities where you’d think traffic was much less of a problem. If traffic is the major issue, it would have to be because helmet laws reduced cycling which led to reduced safety in numbers leading to worse traffic, and an even greater deterrent to cycling.

          • Alan Davies says:

            I think that would be a 27% increase on the number who rode in the past month, not 127%.

          • I’m puzzled by Alan’s maths. At total of 16.05 out of every 100 people who didn’t ride in the past month said they would more except for helmet laws. If these people ride at least once a month, the number of people riding at least once a month would be 16.05+17.8= 33.8, an 90% increase on number of people currently cycling at least once a month.

            On top of that, 37% (6.6 people per 100) who already cycle at least once a month say they would ride more. So the total number who would ride more is 6.6+16.05 = 22.6. 22.6/17.8 is 1.27 or 127% of the number who currently rode more at least once a month.

            of course, Prof Rissel’s survey is only one aspect of the picture. Perhaps the best indication of the true effect of helmet laws is the number of people who said in post-law surveys that they gave up or reduced their cycling because of helmet laws, together with all the propaganda making people think that cycling was so dangerous that no sane person would ride without a helmet. That, together with the effect of reduced safety in numbers, must have put a whole lot more people off cycling.

          • Alan Davies says:

            At 12.28pm you said “The total number who would ride more is therefore 6.58 + 16.05 = 22.64 people, or 127% of the number who rode in the past month”. And the number who rode in the last month, you say, is 17.8. The increase is thus 22.64 – 17.8 = 4.84. That’s an increase of 27%.

            Hmmmmm………that’s not right is it?

        • And by the same logic, if every single one of the 17.8% people who rode in the past month said they would ride more if they wore a helmet, the increase according to your maths would be 17.8 – 17.8 = zero percent!!

          • Alan Davies says:

            It’s doing my head in! Are you adding the 23.4 to the 17.8? I suspect you must be, but the 17.8 are existing regular cyclists, they’re not all saying they’d cycle more if helmets weren’t mandatory. Only 23.64 (6.85 + 16.05) are saying they’d cycle more.

        • It’s easiest to think about this in 2 stages. First the 16.05 potentially new cyclists, in addition to the 17.8 existing ones who ride at least once a month That makes a total of 33.9 new+old cyclists, an increase of 16.05 (90%) on the original total.

          Then you have to add in the effect of the 37% of cyclists who ride at least once a month who say they would ride more. Whatever you believe the effect to be, it’s bound to be more than the 90% increase you get if no-one who cycles at least once a month would cycle more without helmet laws.

          • Alan Davies says:

            That explains your position much more clearly. I will (magnanimously!) concede I misunderstood what you were getting at….but….perhaps it also clarifies why we’ve not seen it the same way.

            You are treating the 16.05 people who say they’d cycle “more” the same as the 17.8 who cycle at least once a month. I concede that’s an “increase” in the literal sense but not in any meaningful sense.e.g. “more” might mean once a year. I think it should properly be described as a ratio, not an “increase”.

            Probably the main source of confusion for me is describing the 37% of regular riders who say they’ll increase their riding as an increase in the number of people cycling. For this group it could only possibly be an increase in the level of cycling e.g kms. And it could be a very modest increase too; we just don’t know.

  10. Nick R says:

    Just a quick addition to the helmet law debate. As a former injury data analyst I have to ask what is the point of all these evaluations of the *frequency* of cycling related head injuries in relation to helmet law? This really tells us very little. What is missing from all of these studies is a measure of head injury *severity*.

    To put this in context with an example: we don’t expect seatbelt wearing to reduce the number of car crashes. Rather we expect the same frequency of car crashes seatbelt or none, and the same numbers of people treated for car crash related injuries… but the severity of the injuries sustained in a crash is greatly reduced if the occupants are wearing a seatbelt. In most cases the occupant still goes to hospital, but seatbelt usage can mean the difference is between a concussion and life in a wheelchair, or between a broken arm and the morgue. Same number of crashes, same number of injured people treated, massive difference in injury outcomes.

    Likewise we shouldn’t expect any less (or more) bicycle crashes with helmet wearing, as helmets aren’t a contributory factor in crashing. Similarly we shouldn’t expect any less head injuries, as in most cycle crash scenarios a strike to the head is still going to result in a head injury. But if cycle helmets work, what we should see is a reduction in the severity of those injuries.

    To put it in real terms if you have a serious crash on a bike you are probably going to end up having medical treatment for a head injury, helmet or no helmet. But the helmet could mean the difference between that head injury being a mild concussion and vegetative coma or death (to be extreme). On paper that means the same number of head injuries in either case, but the real life outcomes are vastly different.

    • There are published analyses of injury severity. For example Robinson (Health Promotion Journal of Australia, 2005) looked at the percentage of injuries involving death or serious head injury (%DSHI) in Transport Accident Commission claims in Victoria. Concussions were not included in the definition of serious head injury.

      Comparing the two years before the helmet law with the two following years) %DSHI for pedestrians fell to 92.7% of its pre-law value, perhaps because campaigns against speeding together with increased random breath testing reduced collision speeds and made drivers more alert to the presence of pedestrians.

      For cyclists, the fall was almost identical – to 93.9% of the pre-law value. So there’s no evidence of any additional benefit over and above what happened to pedestrians.

      I’m surprised at your argument that we shouldn’t expect any more crashes. UK researcher, Dr Ian Walker, measured the distance left by overtaking motorists, and found that he was given significantly less room when he wore a helmet. He was hit twice when conducting this research – by a truck and a bus – both times when he was wearing a helmet

      In another study, Norwegian researchers measured downhill cycling speed of volunteers who were used to wearing helmets. They cycled faster when wearing their helmets than without. This could also translate into increased crashes, if only because vehicle drivers are prone to under-estimate cycling speeds.

      In Australia, only 2 states counted cyclists at the same sites, observation times and the same time of year before and after helmet laws. In Victoria, all cyclists were counted. The fall in TAC claims for serious head injuries was less than expected from the change in cycling and the change in pedestrian injuries. In NSW, there was no pre-law count for adults at the same time of year as the pre-law count, so there’s no reliable information on the change in adult cycling. But for children, the fall in hospital admissions for head injury (the usual definition of serious head injury) was less than the reduction in counts of children cycling.

    • The main arguments against the helmet law are essentially that MHL suppresses bike riding, which means fewer bikes on the road, which has the follow on affect of meaning drivers are less aware of bikes, resulting in more crashes. In short the “safety in numbers” principal is reduced.

      Furthermore a few studies have shown that drivers tend to give cyclists with helmets less room when overtaking which can contribute to an increase in accidents.

      • Alan Davies says:

        The critical issue is by how much the “MHL suppresses bike riding”. I suspect it does to some degree, but I’m not sure the effect is anywhere near as significant as the anti-MHL lobby contend.

        There’s also a possible counter-effect i.e. parents who buy their kids bikes at younger ages – or let/encourage them to cycle more – because the parents feel the MHL increases the likelihood their children will wear a helmet.

  11. […] Other criticisms of the Rissel and Wen study by Dr Alan Davies can be found on his Melbourne Urbanist blog. […]

  12. Tim Churches says:


    You and your readers may find this response to the claims made by Chris Rissel based on his most recent helmet research of interest:

    Tim Churches

    • Tim wrote: “In order to substantially increase cycling participation and mode-share … Australian governments and authorities need to focus on strategies which have been proven to work in Europe.”

      Strategies like City Bike Schemes? Consider how they boosted the popularity of cycling in Paris, London and various other cities, but are a total fiasco in Brisbane and Melbourne thanks to helmet laws.

      The striking thing about cycling in Europe is that everyone rides – male, female, thin, fat, old, young. It’s a normal activity done in normal clothes, often as a replacement for short car trips. Although cycling rates vary, I’m not aware of examples where cycling has become popular after helmet laws were introduced. Some UK surveys show an inverse relationship – cities with lower helmet wearing rates had more cycling.

      In WA, trips for education have fallen from about 82 per 1000 population pre-law in 1986 to about 18 in 2006; shopping trips from about 27 per 1000 population to 3. “Other” trips have fallen from about 51 to 23 per 1000 pop. The only cycling that hasn’t fallen dramatically is work trips, which fell from about 15 per 1000 population to 10.

      The measures Tim suggests may work in countries without helmet laws, but there’s no evidence they will work here, any more than the City Bike Schemes. I suspect speed limits will not be reduced without popular support, which is likely to come only if more people cycle.

      Israel and Mexico City were smart enough to repeal their helmet laws to avoid a major barrier to the success of their city bike schemes. It makes more sense than Tim’s argument that if 37% of people who cycled at least once a month say they will ride more without helmet laws, as do 20% of the 82 people out of 100 who haven’t cycled in the past month, that the maximum increase you can expect is 32%!

      A more realistic estimate of the potential is to think of whether we could revert to pre-law cycling rates in WA – education (82 instead of 18 per 1000 pop), shopping (27 per 1000 pop instead of 3) or other trips (51 instead of 23 per 1000 pop).

  13. John Burke says:

    There is another critical MHL issue, related to suppression of cycling numbers as it is but is poignant because it advances an absurd proposition into the realm of percieved reality. Namely that bicycles are dangerous. Well perhaps a little but probably not as much as TV in the longterm.
    Cars are dangerous not bikes and how could we have come to a general consensus that the opposite is so? MHL is a major factor. I think the idea that bikes are dangerous, beyond a certain point, would be laughable if it were not so dangerous itself.

  14. Quora says:

    Are you safer riding a bicycle with or without a helmet?…

    If we’re going to discuss flawed studies (and I’m not admitting the two studies are in anyway flawed)… The Rissel research you linked to on Croaky has been dismantled here  and here  http://melbourneurbanist.word

  15. and for xmas here is a very thoughtful video on this topic

  16. Statistics show that humour is an effective medium to convey a message.
    merry xmas 2 u Alan and all of your excellent contributors to The Melbourne Urbanist who have provided me with some balanced thought with regards to my own extremes.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s