Is Cycling Dangerous?
The principal theme of NormalizeCycling is that the risks of cycling in Britain are much lower than generally perceived. Does this mean that fears are entirely groundless? Can parents let teenaged offspring go free on their bikes? Could everyone jump on a bike and just get on with it?
This page explores the origins of the belief that cycling is dangerous, and examines the nature of “risk” in more detail. It is not based on a peer-reviewed paper. I have drawn on factual material as much as possible, and observations of the rise of fear in cycling since the 1980s.
My conclusions are that:
- There are valid causes for concern when cycling on the current British road system. However, individuals can be taught to minimise known hazards.
- Cyclists are more prone to “spills” and injuries than pedestrians or drivers. These injuries are not usually serious, let alone fatal.
- Changing car design has led to an increase in perceived, and in some cases actual, danger.
- Incompetent risk assessment by road safety institutions and the promotion of helmets have likely intensified trends that would have occurred anyway.
Valid Causes for Concern
On the Risk in Cycling page (see Chart 3), data from the British Department for Transport (DfT) illustrate how traffic speed affects risk for cyclists. A rough rule of thumb is that a 10 mph increase in speed limit doubles the risk of fatality. This means that a cyclist on a short trip could be exposed to prosaic risks for most of the distance, but briefly face the risks of motorcycle riding for a stretch of trunk road.
Similarly, large roundabouts and gyratories are not safe places for cyclists. When I lived in London in the early 1990s, I would not even attempt to ride around Hyde Park Corner, Russell Square, or Trafalgar Square. These were obstacles that I walked around, making my trip longer and more hassle (but it was still better than The Tube in rush hour).
Aside from these specific points of risk, cycling on Britain’s roads is a work-around. Britain’s roads were not designed with cycling in mind. People still cycle because of the determination of enthusiasts to maintain the tradition; because it is just such a good way of getting about. But it is a work-around. This means the individual has to develop their own solutions to parked cars, buses and other large vehicles, pot holes, steel covers, pinch points, slip roads of dual carriageways, anti-social driving and all the other challenges arising from a system designed for motor vehicles. Any one of these challenges can lead to at least a fright if not handled properly. Help is available from training (Bikeability) and the excellent manual Cyclecraft, but these require commitment. For most people, it is easier just to stick with the car. Additionally, one needs to be fairly fit to ride in traffic. Considering all the barriers to entry, it is not hard to understand why cycling remains marginalised in the absence of re-engineering the roads.
The above combination of actual hazard points, combined with the difficulties faced by any beginner cyclist, in my view go a long way to explaining the fertile ground from which public fears have grown. These fears do have a substantive basis. With training, re-engineering of large junctions, and separation on trunk roads, much can be done to alleviate them.
Injuries in Cycling
The study of risk in cycling on this site has focused on the risk of fatal injury. This is because deaths in traffic crashes and falls are reliably recorded in police reports and mortality statistics. Lesser injuries are not so reliably recorded. Notwithstanding this, there is evidence that cyclists are more prone to injury in traffic collisions than pedestrians or drivers.
In Britain there are typically 100 cyclist deaths per year in crashes. In addition, there are around 3,400 serious injuries (requiring admission to hospital or involving various significant injuries, including death more than 30 days after the incident). Whereas, there are around 450 pedestrian deaths, but the number of serious injuries recorded, at just over 5,100, is only a bit higher than the figure for cyclists. [Note, these figures are from STATS19 police reports, which focus on injuries related to motor vehicle use. The figures for cycling include fatalities due to falls in the highway (typically 10-15 per year) and a small number of serious injuries in falls. The figures for pedestrians do not include any casualties due to falls. However, the inclusion of falls for cyclists does not alter the gist of the general observation being made here].
Now, I have good reason to believe that pedestrian injuries are under-reported by the police. I am in the process of drafting a short paper on this topic, based on comparing police reports with hospital discharge records. Nonetheless, even taking hospital records, the number of seriously injured pedestrians is low relative to cycling. Pro rata with cyclist deaths and serious injuries, one would expect 15,300 pedestrians seriously injured by vehicles per year. In fact, hospital records show less than half this number.
Moving down the hierarchy of severity, the next class is “slight” injury. This means that the emergency services attended an incident involving at least one motor vehicle and someone was treated for injuries not classed as “serious” or later recorded as “fatal”. [Note: the figures for cyclists may include some falls in the highway]. “Slight” injury is therefore a broad-brush category. It does not mean “trivial”. There are plenty of unpleasant sprains, bruises, gashes or lesser bone fractures within the “slight” category. The police record 15,000 slightly injured cyclists and 18,000 slightly injured pedestrians. This repeats the pattern of a higher rate of lesser injuries for cyclists than pedestrians, although there is no way of knowing how complete these figures are.
The “slight” injury category is not very meaningful. The great majority of injuries when walking or cycling do not get recorded anywhere in a systematic way.
To be anecdotal, I have suffered injuries falling off a bike on three occasions in the last twenty years, yet none required attendance at an A&E department. I have also been injured falling on steps, but again, nothing worse than a few scrapes and embarrassment. I have occasionally witnessed a cyclist knocked off their bike, without the police or an ambulance subsequently attending the scene. Officially, these incidents do not exist. That does not mean the incidents lacked an element of fright.
Had I attended A&E for treatment, this would have been recorded. My injuries would have acquired official existence. Unfortunately, there is no national record of A&E attendances. These data used to be collated from a sample of general hospitals around the country in the form of HASS and LASS reports, but the series were discontinued after 2002.
One final source of information has appeared recently as part of the National Travel Survey. The survey now asks about travel injuries (of any severity) that happened to the participant. The system has the weakness of being self-reported. Analysis shows that incidents (as self-reported) are commoner for cyclists than pedestrians. [Note, these injuries could have been related to either a collision with a motor vehicle or a fall].
The overall pattern supports the conclusion that cycling incurs a higher rate of non-fatal injury than walking. To what extent this is due to many British cyclists being sporting riders – much more injury prone than utility riders – is not clear. Higher injury rates provide additional fertile ground from which fears about safety may germinate.
Changing Car Design
I have already made some observations about how larger, more powerful and agile cars have, in my view, contributed to an increase in perceived, and in some situations actual, danger. See my page on Safer Cycling Feels More Dangerous?
Incompetent Risk Assessment
Prior to the development of the National Travel Survey (NTS), the only source of information about risk on the roads came from vehicle counts for the National Road Traffic Survey (NRTS). The NRTS did not include pedestrian counts. Its results showed that per KM of travel, cycling was more than ten times as risky as driving. I further discuss this traditionalist view of risk on this page about Units of Risk. The absence of information about pedestrian risk left cycling classed with motorcycling as an (apparently) high-risk form of travel.
A further problem arose from efforts to verify police reports against hospital records. These comparisons apparently showed a much larger difference for cyclists than pedestrians. This gave rise to the widely-stated official view that “cyclist casualties are the most under-reported of all road users”. This view is wrong.
I will minimise the nitty-gritty minutiae of the causes of the problem. Basically, cyclists admitted to hospital are recorded (in hospital records) in a different way to pedestrians. All cyclist admissions are classed as “transport injuries”, including children injured at play or mountain bikers injured riding off-road. Whereas, only pedestrians hit by vehicles are recorded under “transport injuries”. This means that pedestrians who fall down steps, or whatever, do not get recorded as transport injuries anywhere, not even in hospital records. Instead, pedestrian falls are included in a separate category within “Other Causes of Injury”. Even in that category, the recording is incomplete, since the “location of injury” code is only available for about 25% of falls.
Now you can appreciate the potential for confusion when hospital records are used by transport bureaucrats or academics who do not understand the classification system. They end up overstating cyclist injuries, whilst omitting altogether the large number of pedestrians seriously injured in falls (in Britain, greater than the number of serious injuries in all traffic collisions). My observation is that vanishingly few researchers are aware of this point even now, notwithstanding that a paper I co-authored many years ago described the matter in detail.
During the last fifteen years or so, the NTS has improved to provide data on distance walked, which can be used to estimate risk in walking. For all countries I am aware of, the risk (of fatality) per KM is higher for pedestrians than cyclists, although the risk per hour may be lower.
The much more detailed risk assessments pioneered by University College London researchers (with a little assistance from me…) have, I hope, clarified the situation and eased the official stereotyping of cycling as “dangerous”. These papers from the basis of the Risk in Cycling page of this site.
Helmet Promotion / Helmet Laws
It is not possible to know to what extent helmet promotion directed only at cyclists (and helmet laws in other countries) have exacerbated fears for the safety of cyclists. Did underlying fears lead to increasing helmet use, or did helmet promotion germinate fears in soil fertile with some actual causes for concern? Was it a reinforcing cycle of fear, as rising helmet use led to anecdotal “scary stories” and media sensationalism, which caused more fear? I simply do not know. I do not know if it can be known. My position – based on established principles of health and safety – is that the true relevance of helmets in road safety policy has been greatly over-stated, for reasons I suggest on the cycle helmets page of this site.