It has been identified that there is a difference between actual safety (ie what is experienced and observed through crash records) and perceived safety (ie how people interpret environmental factors which affects their willingness to cycle).  This section summarises the main actual safety concerns for people who cycle and is largely based on 2008–12 data in the Crash Analysis System (CAS) maintained by the NZ Transport Agency. Until recently police did not report cyclist-only crashes to CAS, so information on them has been drawn from hospital admissions and ACC data. Figures in brackets below and in charts are deaths and serious injuries in five years unless otherwise labelled.  

  • Road type

    There are significant differences in the types of risks for people cycling on rural and urban roads.

    Note: the data for kilometres cycled is estimated from digitised trips in the Ministry of Transport household travel survey thus it is subject to some error; regardless, this is a useful indicative chart.

    The importance of speed can be seen in the following chart of the effect of impact speed on the risk of death or serious injury. While this chart is derived from international pedestrian research, the cyclist data is very similar.  


    Close Back to top
  • Crash severity

    Mackie and Scott conducted a comprehensive review of the studies on the relationship between speed of impact and crash severity for NZ Transport Agency and developed the above summary chart for pedestrians struck by light vehicles. Research by Mackie demonstrates that the severity curve for cyclist and pedestrians is very similar. The bands are broad because the severity of injury in a pedestrian or bicycle crash is not just a function of collision speed, but the age of the person and the type of vehicle striking them also have major influences. Older people are much more fragile and suffer serious injuries or death at lower crash forces than young people. 

    An impact with a heavy vehicle can be fatal at very low speeds. Vans and high-fronted SUVs are significantly less safe than passenger cars with front bonnets. The charts below indicate that to achieve a safe system for cycling, it would generally be necessary to limit impact speeds to below 30 km/h (or even lower if older cyclists are to be accounted for) and minimise potential for conflict with heavy vehicles. The relative risk of death if struck by a passenger car driving at 50 km/h is four times that at 30 km/h.  The risk of death if hit at 60 km/h is eight times that at 30km/h.

    Close Back to top
  • Rural cycling crashes

    On rural roads, four out of five cyclist deaths and two-thirds of all crashes resulting in death or serious injury do not involve intersections or driveways. The ‘typical’ crash in this group involves a cyclist being struck from behind on a straight road. In most of these crashes the cyclist was not seen in time and some were overtaken with insufficient clearance, giving a high speed of impact. One in six crashes in this group occurred at night; given that cycle volumes are night are much lower than during the day, this suggests a high risk for night-time cycling.

    Lack of shoulder width is a significant factor in rural cycling crashes.  Rural cyclist crashes were plotted on a map and the sealed shoulder width compared to records and aerial photos. Crashes are scattered widely across the network and the majority happen where there is no road shoulder (see figure below). The lack of shoulder width results in a greater risk to cyclists of being hit by heavy vehicles which, being wider, require more room to safely overtake.

    There were also a small but significant number of children killed riding out from their own farm driveway entrance. Other rural crashes are ‘failed to give way’ conflicts with similar issues to the urban problems described next.  

    Close Back to top
  • Urban cycling crashes

    An urban road is one with a speed limit of 70 km/h or less. From 2008–12 there were 19 cycling deaths and 748 serious injuries on urban roads reported by the New Zealand Police. 

    Intersections and driveways present by far the greatest urban risk, accounting for three-quarters of crashes resulting in death or serious injury. Where one party was required to give way, five times out of six it was the driver of the motor vehicle who failed to do so. Almost every time the driver claimed that they did not see the cyclist they were required to give way to. This phenomenon, referred to as ‘sorry mate, I didn’t see you’, was by far the dominant factor for urban cycling injuries, present in 312 of the crashes in this sample. Fortunately, in most of these crashes the motor vehicle speed was below the 30 km/h threshold above which death or serious injury is most probable. However, where heavy vehicles are involved, consequences are more severe.  39% of urban cyclist deaths involved a heavy vehicle.

    Priority-controlled (ie give way or stop) ‘T’ and ‘X’ junctions are by far the most common on the network so not surprisingly these intersections have the highest number of cyclist collisions (309 serious and four deaths). They typically involve motor vehicles emerging from a side road, or turning right across a cyclist’s path from the opposite direction. Often the cyclist has been riding past stationary traffic and the driver has turned through a gap. Research indicates that the lowest risk to people cycling through intersections is when they are positioned near the centre of the main traffic lane (where they will be in the central vision of a driver looking for another vehicle). The same principle applies at driveways and roundabouts.

    Roundabouts have the highest risk per cyclist, because drivers who are required to give way before entering roundabouts travel faster on the approach than at other intersection types. Drivers tend to check for conflicting traffic further away on the approach, and thus look past cyclists who are often positioned in the periphery of their vision.  Larger and multi-lane roundabouts are the worst. Roundabouts are the safest form of intersection for motor traffic, because they reduce impact speeds to below the safe system thresholds for serious injury; however the operating speeds on roundabouts are typically too high for pedestrians, cyclists and motorcyclists under safe system principles, so a high proportion of injuries at roundabouts involve cyclists. Roundabouts were involved in 68 serious injuries and four deaths in the urban cycling crash sample; three of the deaths involved trucks.

    Sixty-five serious injury crashes and two deaths in the urban cycling crash sample occurred at intersections controlled by traffic signals. At signalised intersections the risk from side road traffic is well controlled but cyclists are still at risk from traffic turning left or right. For left-turning traffic, research shows that where the left lane is marked for both through and left turning traffic, the risk is about four times that where there is a left turn only lane marked. Layouts with multiple through lanes where it is permitted to pick a gap to turn on a full green signal pose the greatest risk from right turning traffic. 

    Driveways were the location of 109 deaths and serious injuries in the urban cycling crash sample; 31 of these (including three deaths) involved people cycling on the footpath. Recurring factors include reversing from the driveway, children riding out from the driveway and drivers failing to see a cyclist when turning into or leaving a driveway.

    Riding on the footpath was identified as a crash cause for 51 crashes in the total urban sample. Most studies show that riding on the footpath involves a greater risk than riding on the road. The main risk is from riding from the footpath to cross a road. However, there are a number of situations where the risk on road is higher due to fast and busy traffic; if there are few busy driveways or side roads it may be safer to cycle on adjacent footpaths in these locations. Such footpaths could be designated as shared paths on a case-by-case basis. 

    Parked cars create a number of hazards for people cycling. It is acknowledged that some crashes (16 crashes in the sample considered) involving parked cars are due to people cycling inattentively with their heads down and thus running into to the back of  a parked car. Others are due to the car drivers or occupants: eg opening car doors into the path of someone cycling (48 crashes in the sample, including two deaths); drivers failing to notice cyclists while manoeuvring into or out of a parking space (58 crashes in the sample) and cyclists moving into the path of overtaking traffic whilst trying to avoid collision with parked cars or similar obstacles (21 crashes in the sample, including one death, which involved a passing truck). 

    Unlike on rural roads, deaths and serious injuries on urban roads due to being struck from behind are rare (18 in this sample). However this crash type is important because the proximity of overtaking traffic is an important contributor to cyclists perceptions of risk on urban roads, and hence the need for increased separation from traffic to overcome the bravery barrier to increase cycling participation.  

    Close Back to top
  • Non-motor vehicle cycling crashes

    Most crashes resulting in injuries to a person cycling do not involve a motor vehicle. They may involve a pedestrian, other people on bikes or the injured cyclist may be the only person involved.

    Until recently, police did not record crashes not involving a motor vehicle. So, while hospital admission records can give an indication, there is little formal information available on crash characteristics and causes. Information reported by the from the Ministry of Transport(external link) shows that in 2016 there were 104 cyclists hospitalised (for over one day) from crashes involving a motor vehicle, an additional 361 cyclists were hospitalised from traffic incidents not involving a motor vehicle and another 493 from non-traffic incidents, (ie off-road riding such as BMX, downhill and mountain biking). 

    The number of cyclists hospitalised from non-motor vehicle crashes is about three times the number hospitalised from collisions with motor vehicles. However, on average the length of admission (number of days stayed) from cycling crashes involving motor vehicles is greater than those not involving non-motor vehicles (6.3 days as opposed to 4.6 days).

    Two New Zealand studies have interviewed people further about their cycling crashes. Munster et al (2001) found that 28% of crashes were attributed by the cyclist to road features. Of these, gravel or debris on the surface contributed to 34%, and surface irregularities (potholes, anything trapping the bicycle wheel, uneven surfaces and judder bars) accounted for 39%.  

    Cycle-only crashes in urban areas mainly involved people cycling for transport, while crashes in 100 km/h speed zones mainly involved people cycling for sports training. Cyclists attributed most of these crashes to their own inattention and speed. 

    In hillier parts of the country, downhill grades are likely to increase the risk of a cycle crash (as found in Vancouver(external link)). This risk should be considered when choosing facility types and locations, and when determining specifics such as facility width, radius of turns, camber and sight lines.

    Close Back to top
  • Safer cycling practices

    The above analysis suggests that the following key cycling competencies for avoiding crashes with motor vehicles:

    • cycle in the correct position – clear of car doors and where most visible to drivers of motor vehicles
    • check behind before merging to the right or turning right
    • stay behind vehicles indicating a left turn, especially trucks
    • keep visible at night – in urban areas being conspicuous from the front is most important.  In rural areas it is most important to be visible from the rear, although the risk still appears high even for people cycling with rear lights
    • overtake queues on the left slowly and look carefully for traffic turning through gaps in the adjacent queue
    • know when to claim the lane.

    An old US study (Kaplan) found that experienced cyclists with the skills outlined above had a crash rate five times lower than university students in general.

    Close Back to top