Keen to ride your bike around town, but worried it’s a risky business? You’re not alone, and that’s why we work for safer streets for people on bikes. But hold up! When you crunch the numbers, even under current conditions, riding a bike is not nearly as dangerous as you might think, compared to other things we get up to.

As reported in the NZ Herald, new research from a team led by Prof. Alistair Woodward at the University of Auckland shows that biking is about as safe as doing DIY around the house. We’re glad to be able to share Alistair’s own blog post on the project (also published at Sciblogs)….

“It is too dangerous” is the reason given most commonly for not riding a bike on the road in New Zealand (Legge and Landtroop, 2013). About half of New Zealand households own one or more bikes in good working condition, and most trips in the city are relatively short, but only 1-2% of those trips are made by bike (Ministry of Transport 2015).

In a study that was published last week we set out to determine how big the risk of injury due to cycle crashes actually is, and how it compares with injury risk of other common activities in New Zealand.

What should these comparisons be based on? Risk per hour, per kilometre, per head of population? We have used the notion of a “standard dose”. We were inspired by Prof David Nutt, an English physician and researcher. Nutt thought the regulation of psychotropic drugs should be in proportion to the damage these drugs caused, and he compared the probability of serious ill-effects per standard dose (what a user would commonly ingest) with the risk of injury during a typical day of horse-riding (Nutt 2009).

We proposed a standard dose of cycling might be a 30 minute trip, 3 times a week. This is rather more than the average cyclist in New Zealand presently spends on his or her bike, but fits with what national and local governments aspire to achieve in the near future (Auckland Transport, for instance, aims to triple the number of journeys by bike in the city in three years).

The comparison activities we selected were snow sports, do-it-yourself (DIY) activities in the home, rugby, horse riding and using quad bikes on farms. These were chosen on the basis that they are common, and there are data available on both person-time exposure, and the rates of injury. We used two injury outcomes. Injuries sufficient to cause the person affected to visit a hospital emergency department, and injuries that lead to a claim being made to the Accident Compensation Corporation (ACC).

We found the risk of injury while riding a bike is actually very small. Taking injuries that lead to claims to ACC, we found these occur roughly 9 times in every 100,000 short urban bike trips; the chance of receiving an injury sufficiently severe to cause a visit to the hospital was similar. If you rode a bike three times a week, most weeks, the chances are you would suffer one moderately severe injury every 70 years.

We estimate the risk of injury on a bike is similar to the risk associated with DIY activities in the home, more than a 100 times less than the risk of snow sports, and 500 times safer than playing rugby (see Figure 1, which uses a log scale).

Figure 1. Risk of injury sufficient to cause a visit to a hospital emergency department, or lead to a claim to ACC, per million typical exposures (such as a half hour bike ride, a game of rugby, half a day on the snow)

There are many assumptions and approximations in these results, so they should not be treated as precise measures. But the take home message, we suggest, is that riding a bike on New Zealand roads is not particularly dangerous (1 moderate injury in 70 years!), and indeed the risk is considerably less than that associated with some other common activities. (Many parents are reluctant to allow their children to ride to sports, but the bike trip is roughly 500 times safer than a game of rugby, for instance).

Risk is a construct – how the figures are presented affects the way they are perceived. The risk of injury per km travelled by cycle on New Zealand roads is higher than that for automobiles but the difference is reduced considerably if automobile casualties include those injured by cars as well as those injured in cars, and if travel on motorways is excluded (given that motorways are low risk environments, with no equivalent for bikes). For some sub-groups (eg, young men) the risk of injury in a car is greater than that when travelling by bike, yet bicycles are not seriously promoted as a means of reducing crash injuries among the young (Mindell et al 2012). The rate of a fatal injury per kilometre is higher for pedestrian travel than either cars or bikes (Shaw et al 2016), yet few people are put off walking by fear of traffic crash injury. Also, think of this – travelling by private motor car is an order of magnitude more dangerous, in terms of crash injuries, than going by public transport, but this is seldom used as an argument for more buses and trains. And why focus just on injury? If deaths from all causes are counted, cycling is a good deal healthier than driving in a car, as premature deaths avoided by extra physical activity (via reducing heart disease and cancer risk etc) exceed by roughly 20 to 1 cycling-related injury deaths (Lindsay et al, 2010).

So why does fear of injury deter so many people from getting around on a bike? We suggest the first point is that the probability of something happening is, on its own, a weak motivator. Statistical risk only makes sense when it is filtered and interpreted through personal perceptions and social frameworks. Daniel Kahneman won a Nobel Prize in Economics for his work on what he called the “psychophysics” of risky choices, which he suggests are conditioned strongly by expectations and framing (Kahneman and Tversky, 1984).

In the same vein, we argue that fear of riding a bike is primarily a consequence of living with a transport system that is dominated in every way by the motor car. The bicycle has literally been pushed to the margins (Figure 2) and the environment sends a powerful message, that such use of the road is unusual, different, and is not valued. The transport norm is reinforced in other ways. For example, cycling promotion campaigns with safety-oriented messages such as “Share the road” have, perhaps unwittingly, strengthened the social framing of cycling as an activity that is inherently dangerous.

Figure 2. Tamaki Drive, Auckland

In the past the hostile conditions on New Zealand roads led to a vicious spiral. Car domination means fewer bikes, leading to greater fearfulness and increased resistance to road changes in favour of bikes. We need to turn this round.

The most powerful way to bring bikes back from the margin is to provide safe spaces for cyclists of all abilities to get to where they want to ride. Separated cycle ways are part of the fix, but not enough. There need to be changes on the road as well, such as slower vehicle speeds, better intersections, and wider shoulders to include cyclists. More people riding, and public spaces that celebrate two wheeled choices, will do two things – make cycling (even) safer, and reduce the fear of the bike.


Kahneman, D., Tversky, A., 1984. Choices, values, and frames. American Psychologist 39, 341-350

Legge N, Landtroop R., 2013. Auckland Transport Cycling Research. Auckland Transport.

Lindsay G, McMillan A, Woodward A, 2010.  Moving urban trips from cars to bicycles: impact on health and emissions. ANZJPH, 35, 54-60.

Mindell, J.S., Leslie, D., Wardlaw, M., 2012. Exposure-based, ‘like-for-like’ assessment of road safety by travel mode using routine health data. PLoS One. 7, e50606.

Ministry of Transport, 2015. 25 years of New Zealand travel: New Zealand household travel 1989-2014. Wellington, New Zealand Government.

Nutt, D.J., 2009. Equasy – An overlooked addiction with implications for the current debate on drug harms. Journal of Psychopharmacology. 23, 3-5.

Shaw C, Russell M, van Sparrentak K, Merrett A, Clegg H, 2016. Benchmarking cycling and walking in six New Zealand cities. Pilot study 2015. New Zealand Centre for Sustainable Cities. University of Otago Wellington.

Share this

7 responses to “Fear’s fair – but how rational is our risk assessment?

  1. Agree that the risk is not that high, but that’s missing the point. Even if they don’t kill you, the constant stream of near misses makes cycling a thoroughly unpleasant way to spend your time. Most people won’t cycle if it requires you to be on edge all the time.

    The remarks about figure 2 are right on the money. People don’t cycle (and for a large part don’t walk either) because when you do the city sends all sorts of signals that you’re doing it wrong.

    Also, don’t trust those statistics:

    – people do their commute to work way more often than playing rugby game.

    – commuting typically happens when the roads are more busy than usual. The same applies to children riding to school. These things happen during the most dangerous part of the day

    – and the biggest one is selection bias. People will cycle more often *when they have safe alternatives*. So your statistic will greatly underestimate the risk people would be exposed to in your average part of Auckland.

  2. There is also another issue that is hard to measure… people will often give a reason to surveyors that is not the real reason.. because they don’t want to seem unreasonable or hypocritical (in thier own mind)… They “self sensor” the answers…

    So, you may not cycle because its possibly dangerous, or you just cant be bothered with the physical activity, or you don’t like the way you look… Which one are you going to say “out loud” in public?

    1. Another thing the surveys will not catch, it that for a large part the reason why people do things or not is subconscious.

      If you cycle on the street in Auckland, you’ll receive a lot of signals from the city that you’re doing it wrong. You’re often sandwiched between parked cars and cars going 60. On shared paths you’re required to give way to side streets, while cars have right of way.

      Compare it to using a spade to cut your sandwich. Maybe someone who really likes spades may point out it will work better if you keep your spade sharp, and that you should probably keep your spade clean. Most people however will intuitively figure out that they should use their spade to dig a hole in the ground, rather than cutting a sandwich.

      Likewise, Auckland streets are not designed for cycling. They’re designed for driving, (and often driving fast). Suggesting someone to cycle to work will likely get you the same reaction as that spade example. For the same reasons.

      1. You make some valid points wsomc. I tend to agree somewhat with your view that cyclists are doing it wrong – but not on purpose. They’re railroaded into these pinch points, sandwiched between cars as a result of poor city planning. Changes are afoot to improve infrastructure as the penny has finally dropped that they can’t keep adding more and more roads as cars will simply clog them up. The greatest invention of the modern age – hamstrung by it’s own masses. Hopefully in 10 years significant progress will have been made and people will again embrace cycling, walking. The trick is to gain their trust in order for this to happen and we’re not there yet. Not even half way.

        1. I didn’t mean they’re actually doing something wrong. But current street design is sending those signals anyway, and that has to be fixed. I We’re making progress, but we have an entire city to fix up.

          The same often goes for walking in a lot of places. Things like missing pedestrian legs, extreme reluctance to build legit pedestrian crossings (see the Ponsonby Road consultation for a recent example) and stupid traffic light phasing (like not automatically giving a green man while all conflicting traffic has red light).

  3. good points. I agree that there is a tendency to give a comfortable, socially acceptable answer in surveys, such as “bikes are not safe”, perhaps without a lot of thought. There probably are other things in people’s minds, such as being embarrassed, or unconfident. But this doesn’t make the perception of safety issue unimportant. It gathers momentum because it is so widely reported.

    And I agree that being stressed by traffic isn’t exactly the same thing as worrying about being knocked off. This is relevant to the assessment of bike infrastructure. In the past, attention has been paid to crash hot-spots, which is partial approach, after the event, and as you say, not necessarily dealing with important aspects of road quality.

    On the stats, we were interested in risk per “dose” (eg a game of rugby or typical bike ride). We didn’t attempt a comparison of the total burden of injury due to each activity. That obviously depends, as you point out, on how much rugby people play, or how many bike trips are made. What stands out here in fact is DIY! Just because that is something an awful lot of people spend time doing, far more than cycling or playing rugby.

    Yes, if you wanted to get really fancy with the risk calculations, then you would take into account time of day, type of road, city, weather conditions and a host of other factors. We just took an average, across the year, and across the country, an approximation, but I think it is still a fair guide.

  4. I cycle commute virtually every day. In 16 years I’ve had 2 accidents to from cars. Not ideal of course but my wife who drives to work has had 2 accidents also.
    There are risks in cycling, but common sense can reduce those risks so they are far outweighed by the benefits.

Comments are closed.