Cycle paths are dangerous where they cross junctions

This wiki page is a draft which many people can edit and which should not be considered to be official Cycling Embassy policy.

Common claims and canards > Objections from cyclists to cycle paths > Cycle paths are unsafe > Cycle paths are dangerous where they cross junctions

Summary of the claim

By separating cyclists from motor vehicles, the risk to cyclists is increased on the occasions where the two modes must interact, such as at junctions and side-roads. Where a cycle track crosses a side-road, cyclists are at particular risk of being hit by vehicles turning into the side-road. This negates any safety benefit of separation.

Example sources

When this claim is made, references are typically made to a Danish study, “Road Safety and the perceived risk of cycle facilities in Copenhagen.”1. This paper is often quoted with the eye-catching figure of cycle tracks resulting in a 24% increase in crashes involving cyclists at intersections where cycle tracks have been implemented.

Summary of responses

  1. There are many different designs of roadside cycle tracks, not all of them equal. The Cycling Embassy actively campaigns against poor quality designs, such as the existing British shared pavements. The high quality cycle tracks that we advocate have been implemented in the Netherlands, with a number of simple and successful solutions to the junctions problem.
    1. Dutch junction geometry, and the use of raised tables for cycle track and footway crossings at side roads, forces motorists to slow to a speed at which they have clear time to see the cycle track and check who is on it.
    2. The normal geometry of Dutch cycle track crossings ensures that motorists and cyclists are at 90 degree angles, meaning that each can clearly see the other.
  2. There are several limitations to the Copenhagen study commonly cited.
    1. The study in fact found a reduction in crashes. The “24% increase” is compared with a predicted crash rate.
    2. The simple before-and-after study design can not separate the effects of interaction between drivers and the cycle track design from, for example, the effect of less experienced bicycle users being attracted by the new infrastructure.

In more detail

Some cyclists have actively campaigned against the introduction of certain forms of cycling infrastructure, arguing that they are more dangerous than cycling on roads. Specifically, they cite the case of cycle tracks running parallel to a road — a type of cycling infrastructure which the Cycling Embassy advocates for use on fast and busy roads, such as ‘A’ roads. The argument has been made that cycle tracks put cyclists out of sight and out of mind of motorists, so that when cyclists and motorists do have to interact — where a cycle track crosses a side-road, for example, or where two main roads intersect — motorists will not be expecting or watching for cyclists crossing their paths. A study of crashes involving cyclists in Copenhagen before and after the introduction of roadside cycle tracks is typically cited as evidence to support this argument.

Although this argument may be true for poorly designed cycle tracks, the Cycling Embassy believe that the junctions problem has been entirely solved by well designed tracks, as implemented widely in the Netherlands. These tracks use a number of engineering features and rules to ensure that all users are aware of each other’s position and trajectory, and who has priority where their paths cross:

  • Dutch junctions typically have far tighter turns than British junctions, forcing motorists to slow to a speed at which they can also assess pedestrian and cycle crossings.
  • In the Dutch system, where there is a main road with minor side roads, the traffic continuing straight on the main road always has priority over traffic turning into or out of side roads. This means that cyclists continuing straight have priority over turning motorists. It also means that pedestrians walking beside the main road have priority over cyclists and motorists turning into or out of the side roads. Cycle tracks and pavements are therefore usually continuous along the main road, with raised tables at side roads, so that motorists stop and then drive up and over the pavement, rather than pedestrians stopping and then walking out across the side road, as is typical in the UK.
  • Where there is room — typically in rural areas — the cycle track is set back from the main road, so that a car turning into the side road has made the full 90 degree turn before they encounter the cycle track crossing, meaning that they have clear sight lines along the cycle track to see approaching cyclists.
  • At junctions between two main roads, the Dutch model continues to protect cyclists in dedicated cycle tracks, using a simple but clever pattern to the cycle of the traffic lights to ensure that there is no conflict between the movements of cyclists and motorists.

The Cycling Embassy wishes to see these tried and tested engineering solutions adopted, and an end to shared use pavement conversions, which really do put cyclists in danger.

Cycle track crossings of minor side roads

Four-wheeled vehicles such as cars tend to have a fairly large turning radius, meaning that junctions between roads need to have a wide angle in order to facilitate turning at speed. This kind of design is presently commonplace in the UK; where a minor and a major road meet, the intersection is usually designed to be as wide as possible in order to facilitate the turn being taken as fast as possible. This has negative consequences for pedestrians where the pavement meets a side road, as it means that they either have to cross a wider stretch of road or go out of their way to cross the road where it narrows again. Putting a separate cycle lane alongside a road with priority over minor roads, whilst the junctions with minor roads are designed in this manner may encourage dangerous behaviour by motorists and increase the risk cyclists are exposed to. However, this undesirable behaviour can be designed out by changing the way in which major and minor roads meet on roads along which there is a separate cycle path. By narrowing the merging point of the minor road with the major road, the angle of the corner is made sharper, forcing motorists to slow down almost to a complete stop before turning across the cycle path, reducing the risk of a collision with a cyclist2. Where a cycle lane crosses a minor road, it should also be elevated, in order to further reduce the speed of motorised traffic entering or exiting the minor road, reducing collisions with cyclists3. This type of design also benefits pedestrians by reducing the distance to be crossed at minor roads and reducing the speed of motor traffic they are likely to encounter whilst doing so.

Cycle tracks at junctions between main roads

Similarly to where cycle paths have priority over minor roads, the risk cyclists are exposed to be designed out of junctions too. This can be done in a number of ways, including; ‘Simultaneous green’ (also known as ‘scramble junctions’) give bikes and pedestrians their own green phase, allow cyclists to go in any direction including diagonally across a junction4, or the ‘jughandle,’ where at a cross roads the cyclist first crosses the side street, then waits for a light to turn across the original road, ending up on a cycle path or lane on the correct side of the new road2. Somewhat less preferable, this avoids having to move out into traffic and change lanes, but the disadvantage of this can mean having to wait for up to two sets of lights to make the turn.

Limitations of the Copenhagen study

The Cycling Embassy seeks the adoption of the Dutch style of cycling infrastructure precisely because its simple but clever rules and engineering innovations have made it the safest and most comfortable in the world. Cycle tracks in Copenhagen do not follow quite the same rules or styles, and therefore the study which is most commonly used to support the claim that cycle tracks are dangerous at intersections with roads is therefore unlikely to be relevant.

It is still worth considering, however, that the Copenhagen study has a number of limitations, and it is unlikely to be the case that even the Copenhagen style of cycle tracks are dangerous to the extent that has been claimed. The Copenhagen study compared crashes involving cyclists before and after the installation of on-street cycle tracks on main roads in the city. This before-and-after design makes it impossible to completely isolate the effects of the changes to the choreographing of cyclist-motorist interactions from, for example, changes in the population of cyclists using the streets. Cycle tracks enable a larger number and greater diversity of people, including many less experienced and able, to take to their bicycles. The only reason those people were not having crashes before is because they did not dare to cycle on the main road.

Further, it is simply not true that the Copenhagen study found a 24% increases in crashes. Prior to installation of cycle tracks, cyclist injuries at junctions had been measured as 353. After the installation of cycle tracks, the number of cyclist injuries at intersections was measured as 285, a reduction of 19% in absolute figures. The 24% increase figure is calculated from a predicted number of crashes figure for the after period, based on the changes to the traffic volume and mode composition, which predicted that at unmodified intersections with the same increase in cyclists, decrease in motorists and subject to pre-existing crash trends seen at the intersections which had been modified with cycle tracks, there should be 230 cyclist crashes. This is the figure which is used to generate the eye-catching 24% increase in crashes figure. Whilst there may have been pre-existing downward trends in the rates of crashes involving cyclists, it is also important to consider that this effect is seen contemporaneously with an increase in cyclists’ mileage of 20% on these facilities.


1 Road Safety and the perceived risk of cycle facilities in Copenhagen

2 Cycle Tracks: Lessons learned

3 Bicycle facilities on road segments and intersections of distributor roads

4 Dutch Cycle Infrastructure