Let’s be real: roundabouts aren’t exactly the sexiest topic in the world (unless you’re really into traffic engineering). But if you’re a cyclist and you ride through roundabouts with any regularity, you’ll want to read on.
To set the scene, it’s worth noting that roundabouts are increasing in number around the world. In the USA alone, the number of roundabouts grew from zero in 1990 to 4,200 by the mid-2010s. And it’s not hard to understand why they’re being built in increasing numbers. Roundabouts improve traffic flow by removing stop signs and traffic lights, they reduce the number of conflict points between road users, and they reduce the speed of vehicles at intersections. As a result, they help to reduce the number of traffic injuries and fatalities.
But while roundabouts offer a lot for traffic flow and motorist safety, they aren’t nearly as good when it comes to cyclist safety.
What we know
A 2008 study of 91 roundabouts in Flanders, Belgium showed that the installation of roundabouts led to a 27% increase in “bicyclist injury collisions” and an increase of more than 40% in the number of fatal or serious injury crashes involving cyclists. Meanwhile, a 2013 study of more than 300 roundabouts in Denmark found that the installation of roundabouts led to a 65% increase in bike crashes and a 40% increase in injuries.
These findings are certainly troubling, but it’s not as simple as saying that roundabouts always make intersections more dangerous, in all locations, in all conditions, all the time. Different jurisdictions feature different roundabout designs, different attitudes towards cyclists, and different volumes of rider crash data, all of which make it hard to make general statements on the subject.
But the latest work from Niranjan Poudel and Patrick A. Singleton, from the department of Civil and Environmental Engineering at Utah State University, might get us closer to a universal understanding of how roundabouts affect cyclist safety. Their literature review considered dozens of papers on the subject and was published in January 2021 in the journal Transport Reviews.
What the research says
Poudel and Singleton narrowed their search to papers that were written in English and published after 1990, and excluded simulation studies or other literature reviews. They ended up with 49 research papers, most of which came from the cycling heartland of northern Europe (e.g. the Netherlands and Denmark) with a smaller number of papers from Australia, New Zealand, the USA and the UK.
Most of the studies (33 of 49) analysed crash data to draw conclusions about the safety of roundabouts. Others used surrogate measures like “near misses” to make assumptions about roundabout safety for cyclists. These surrogate measures aren’t perfect – the number of near misses, say, doesn’t always correlate with the number of crashes and/or injuries – but in the absence of hard data, they can be a useful back-up.
After reviewing all of the papers, the researchers found they were able to draw some meaningful conclusions about the characteristics of roundabouts, and the types of motorist and cyclist behaviours that can impact rider safety.
They divided their findings into three categories:
How the roundabouts are built and used
A handful of papers found that the higher the number of motor vehicles using a roundabout, and the higher the number of cyclists, the higher the frequency of crashes involving a cyclist. (Interestingly, a couple of papers found the opposite when it comes to cyclists – that there’s a “safety in numbers” effect in play. As noted, a raft of differences between jurisdictions makes things tricky).
Higher speed limits through roundabouts seem to be bad for cyclists too. They lead to a higher frequency of crashes, and those crashes tend to be more severe. As one paper highlighted, slower roundabout speeds meant drivers seemed to yield to cyclists more often.
There’s an interesting wrinkle when it comes to speed though. In 2013 Danish researchers showed that at higher-speed locations, installing a roundabout decreased the number of crashes involving cyclists. At slower intersections though, roundabouts increased the number of crashes by around 100%.
Multi-lane roundabouts, meanwhile, seem to lead to more frequent and more severe crashes, possibly because they allow for greater speeds and more potential conflict points for riders. It’s perhaps unsurprising that cyclists generally perceive these multi-lane roundabouts as more dangerous, more uncomfortable to ride, and as something to be avoided.
On the flipside, roundabouts with larger-radius central islands, or central islands that are higher, or that have greater apron widths, might be safer for cyclists. Poudel and Singleton suggest this might be because larger islands force motorists to considerably reduce their speed, and to focus their attention more carefully when entering the roundabout.
Available bike infrastructure
As you might expect, the type of cycling infrastructure available through a roundabout can have a significant impact on cyclist safety. The authors talk about four common setups:
- No facilities – riders are in mixed traffic.
- On-road bike lane within the roundabout – this might be designated by coloured bitumen and/or on-road markings. Of note, this option is not recommended by most design authorities, but it certainly still exists.
- Shared-use off-road path: Cyclists are funnelled off the road and onto a lane shared with pedestrians.
- Separated bike path: An off-road lane specifically for cyclist use.
One of Poudel and Singleton’s clearest findings comes at this point in the paper. In their words, “research results are consistent” in showing that on-roadway bike lanes within roundabouts lead to “adverse bicycle safety impacts”. What does that mean exactly? Well as an example, Danish researchers found in 2013 that marked lanes within a roundabout increased bike crashes by 33% compared to a roundabout with no bike facilities at all. They also found that a separated cycle path reduced crashes by 84% compared to roundabouts with no bike facilities.
There are clear parallels to be drawn here with painted bike lanes more generally — they might give the appearance of safety, but in reality they can often be less safe than no infrastructure at all.
Interestingly, many riders seem to view roundabouts with painted bike lanes as riskier than roundabouts without any lanes. The authors speculate that the presence of a bike lane might suggest to cyclists that the intersection is inherently dangerous and that they might be at risk.
That’s not a universal finding though — Danish researchers found in 2008 that local cyclists perceived roundabouts with no bike lanes to be more dangerous than those with no infrastructure. This is perhaps just another example of the cultural differences that make it hard to offer universal statements about cycling infrastructure.
Driver and cyclist behaviour
Finally, the researchers looked at the way driver and rider behaviour influenced the safety of roundabouts for cyclists.
They note that where a shared path (cyclists and pedestrians together) or a separated bike path is available at a roundabout, most riders will use it. In situations where there are no cycling facilities available though (either on-road or off-), riders seemed to behave in a handful of different ways.
Some got off the road and instead rode on the footpath, seemingly viewing that as the safer option. Others opted to stay on road, either “straight-lining” it (going from close to the kerb to close to the central island and back to the kerb again) or riding around the outer edge of the roundabout, effectively creating their own bike lane.
After reviewing the literature Poudel and Singleton explain that the most dangerous part of cycling through a roundabout is when the rider is in the intersection, and motorists are either entering or exiting the roundabout. Somewhere between 67-82% of cyclist-motorist crashes that occur in roundabouts are these so-called “entering-circulating” crashes.
Every study has its limitations and this is no exception. Limited availability of relevant crash data is a major obstacle. Many roundabouts are still relatively new too, making it difficult to do before-and-after investigations of the same intersection. Then there’s the fact cyclist numbers are generally rather low (certainly compared to cars in most jurisdictions) which limits the available data. Compounding that issue is a lack of cyclist crash data and the fact that bike crashes are generally underreported.
The researchers explain that more studies are required to help paint a clearer picture of the issues facing cyclists at roundabouts. They even pose some possible lines of inquiry, including: the effect of different roundabout designs; the effect of different lighting and weather conditions; whether socio-demographic issues could be at play; and whether land use or built environment near roundabouts has any impact.
But while more work is required in this space, Poudel and Singleton don’t leave us empty-handed.
What we know now
To quote the authors: “Roundabouts do not improve safety for cyclists as much as for drivers, and may actually increase bicycle crashes, especially for roundabouts with on-roadway bike lanes or no bicycle facilities.”
They add that roundabouts are safer for cyclists when they:
- have a low volume of motor vehicle traffic
- encourage low traffic speeds
- only have one lane
- are smaller in total size, with larger and higher central islands.
As ever in the road safety space, separated bike lanes away from motorised traffic seem to be the most effective way of keeping cyclists safe. The authors note that such infrastructure is particularly valuable in the case of roundabouts with higher traffic speeds and volumes, where there are multiple lanes through a roundabout, where there are lots of cyclists, and where local authorities are keen to encourage so-called “interested but concerned cyclists” – riders who would ride more if the conditions made them feel safer.
As noted above, the most critical concern at roundabouts with no separated bike facilities is entering-circulating crashes. The “look but fail to see” phenomenon is all too common – motorists are used to looking out for other motor vehicles before entering or exiting the roundabout; they’re less concerned with (and/or accustomed to) keeping an eye out for cyclists.
So what does all of this mean for cyclists? If a roundabout offers separated infrastructure, it’s probably safest to use that. If there are painted lanes or no cycling infrastructure, your best bet is to take the lane. That’s certainly the advice of cycling organisations like British Cycling (see video above). Taking the lane should provide your best chance of being spotted by motorists as they enter and exit the roundabout, while also discouraging them from squeezing past you as they navigate the intersection.
And, as ever on the road, it’s worth assuming that motorists haven’t spotted you as you enter a roundabout, and to ride accordingly.