Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Except for one study reporting decreased cycling by children aged 12–17 in Victoria, Australia, Macpherson et al stated: “no other study about the relationship between legislation and exposure to cycling has been published”.1 Macpherson’s paper has already been cited several times as evidence that helmet laws do not discourage cycling, so it is important to examine the evidence.
In fact, several published studies of cycle use and helmet laws are cited and described in refereed journals. Some were attitude surveys. The equivalent of 64% of adult cyclists in Western Australia said they’d ride more except for the helmet law.2,3 In New South Wales (NSW) 51% of schoolchildren owning bikes, who hadn’t cycled in the past week, cited helmet restrictions, substantially more than other reasons, including safety (18%) and parents (20%).4 Street surveys by cyclists’ organisations in the Australian Capital Territory and Northern Territory found 28%–42% of cyclists would ride less/had ridden less because of helmet laws.3
Large scale roadside counts were also conducted. Pre-law surveys counted 6072 child cyclists in NSW,3,5 3121 cyclists (all ages) in Victoria;3,6 and over 200 000 cyclist movements on two key routes in Western Australia.3 Equivalent counts a year after enforced helmet laws showed declines of 36% (NSW), 36% (Victoria) and 20% (Western Australia).3,5,6 Sunday recreational cycling in Western Australia (24 932 cyclists pre-law) dropped by 38%.7 Increases in numbers wearing helmets, 1019 (NSW) and 297 (Victoria) were substantially less than declines in numbers counted (2215 and 1110).3 These surveys were at the same time of year, in similar weather, and used the same sites and observation periods, so can be considered reliable indicators that enforced helmet laws discourage cycling. Reduced cycling activity is associated with increased injury rates per cyclist.8 Australian data suggest injury rates per cyclist increased after helmet laws3; the consequences of reduced cycling because of enforced helmet laws should therefore be fully investigated.
Non-enforced laws may have a lot less impact. One state (Queensland), delayed enforcement for 18 months.9 Helmet wearing increased initially, but 17 months after legislation, rates were little different from pre-law (primary schoolchildren 72%; secondary schoolchildren 21%, commuter cyclists 37%, recreational cyclists 22%, compared with 59%, 13%, 21%, and 22% pre-law).9 Non-enforced laws can be widely flouted, conveying the undesirable message that road safety rules need not be taken seriously, perhaps encouraging dangerous riding and causing more injuries per cyclist than with no law. Enforcing laws unpopular with cyclists may require substantial effort. Queensland issued nearly 23 000 bicycle helmet offence notices annually, 6.7% of all traffic offence notices.9 Per km, cyclists were three times more likely to receive a notice for not wearing a helmet than other road users for all other offences—speeding, drink-driving, failure to stop, give way, or obey other regulations. It is debatable whether this represents the best possible use of police road patrol time.
The Ontario law was not enforced10 and the effect on helmet wearing rates was not reported.1 The results should be interpreted with caution for three reasons. First, the study excluded 15–17 year olds, perhaps the most rebellious age group, despite Australian findings that teenagers were most affected by helmet laws (48% reduction for 12–17 year olds), then adults (29%) with the smallest effect on children under 12 (10%).6 Indeed, in Ontario, only 16–17 year olds could be given direct penalties for non-wearing.11
Second, the annual sample size was much smaller than the Australian studies, only 1128 cyclists observed in 1999 (not 747 as reported,1 for 112 hours of observation; A K Macpherson, personal communication).
Third, the Australian surveys reported weather conditions and used the same sites, time of year and observation periods, to ensure variation due to these factors was not confused with effects of the law. The Canadian study had 111 pre-selected sites, each recorded for one hour, but weather conditions were not reported1 (though elsewhere 1999 was described as a particularly sunny summer; A K Macpherson, personal communication). Table 1 in the Macpherson et al paper shows that, in some years, some sites were recorded more than once.1 Moreover, observations were not at the same time of day and day of the week each year (A K Macpherson, personal communication). Analysis of such data requires careful scrutiny of counts and helmet wearing rates to estimate and allow for differences in observation sites and times, as well as weather conditions, to avoid confusing them with year-to-year variation and the effects of the law. This does not appear to have been done.
Another Canadian survey found “large variations in the number of cyclists per unit time, depending on the time of day, the day of the week, or the month”.12 Perhaps all we should conclude is that the Ontario survey was relatively small, with substantial year-to-year variation, possibly related to year-to-year differences in sites, observations time or weather conditions, these differences swamping any effects of the non-enforced helmet law.
Reply to Robinson
Our study suggested that the enactment of bicycle helmet legislation is not necessarily followed by a reduction in bicycling.1 Robinson cites six observational studies from Australia published largely in reports that do not appear in the peer reviewed literature2–7 contesting our conclusion. Establishing the association between helmet laws and cycling is important, not only in relation to “the consequences of reduced cycling because of enforced helmet laws”, but also because of the possibly substantial effect of helmet laws (enforced and non-enforced) on helmet use rates and head injuries.
The evidence depicting an association between helmet laws and helmet use is strong. All 15 studies that examine this relationship report a positive association.8–23 One study, conducted in Ontario, suggests that the legislative effect on helmet use was greatest among children living in low income areas.8 Evidence related to the association between helmet laws and reductions in head injury rates is also consistent. All six studies that examine this relationship found a reduction in head injuries post-law.11,16,19,21–24
The results of our study on children’s exposure to cycling demonstrated that helmet legislation is not necessarily followed by a reduction in bicycling. Robinson suggests that all one can conclude from our six years of observations is that the “Ontario survey was relatively small, with substantial year-to-year variation, possibly related to year-to-year differences in sites, observation time or weather conditions, these differences swamping any effects of the non-enforced helmet law”. However, our observations were conducted at the same sites, at the same time of the year, and in similar weather conditions. Even if only data from the school observations are included (gathered at the same time of day and the same month each year), there was no reduction in the number of children bicycling to school post-legislation. Further, the helmet law was only “enforced” by the schools’ insistence on children wearing helmets to comply with the law.
The population based impact of helmet laws remains in need of further study. Although evidence for the benefits of legislation is impressive, we agree that it is important to further assess the influence on bicycling. In particular, we believe that evidence is lacking in the following areas:
The effectiveness and effects of enforced v non-enforced laws.
The effectiveness of legislation related to children only v all ages.
The relationship between helmet laws and bicycling among children who have grown up with helmet laws.
Social and cultural differences in the reaction to helmet laws between communities and between countries.
The effect of other trends (for example, the popularity of in-line skating, skateboarding, and scooters) on the use of bicycles.
Characteristics (including bicycling frequency) of cyclists who may be deterred from riding by helmet laws.
The effectiveness of other countermeasures (for example, bicycle friendly environments) to increase bicycling.
Given the benefits of bicycling and the importance of head injuries as a public health problem, it is essential that public policy and legislation is guided by the best available evidence. We need to understand whether, as suggested in our study, bicycling increases after a law because of the positive publicity related to bicycling. If bicyclists in Australia or elsewhere are deterred from bicycling for substantial periods, we need to explore other measures to encourage them to continue. It would help to know which cyclists change their behaviour after the enactment of a helmet law, and if they reduce their bicycling, whether they adopt other forms of activity in place. Finally, the most effective way to enforce helmet laws has yet to be determined.
Given the evidence for the relationship between helmet laws and the reduction in head injuries, and the paucity of well designed studies, perhaps all we can conclude from the studies published to date is that more evidence is needed to resolve the debate.