Friday 21 May 2010

Just how safe are Dutch cyclists and Dutch cycle paths ?

Dutch cycle-path. Four metres wide
and well away from the road. In this
case used by school children.
I've pointed out before that the Netherlands has the safest roads in the world, and that cycle paths have also lead to this country having the safest cyclists in the world. However, for some this evidence simply isn't enough. In the English speaking world in particular, "experts" continue to insist on the ludicrous idea that cyclists are safer in the proximity of cars than without them. However, these countries also continue to have a stunningly low cycling rate.

Now the Germans have joined in too !

The Fietsberaad reports that...

Dutch cycle-path. Four metres wide
and taking a completely different
route to the road. It's good enough
for racing cyclists and it's good enough
for you too.
According to ADFC, the German Cyclists’ Union, people should cycle where they are safest: on the road among cars. Not on separate bike paths, as this would be more dangerous according to various studies. In the United States this news is eagerly distributed by opponents of bike paths. Dutch experts obviously refute this report.
The main argument in the German report is that cycling in the Netherlands is dangerous as 40% of all traffic accidents involve cyclists, whereas these only account for 27% of all travel. And this in spite of the number of bike paths. In Germany, on the other hand, there are fewer bike paths, as well as lower numbers of bicycle accidents in relation to the percentage of travel by bicycle. Conclusion: bike paths are dangerous. This conclusion about the (supposed) danger of separate bike paths in the Netherlands is however by no means well-founded or warranted.

As a matter of fact in the Netherlands cyclists are victims of traffic accidents in not 40% but even 50% of all cases, data provided by Fietsberaad demonstrate. But the percentages and statistics of CBS/Ministry of Transport, Public Works and Water Managements do not at first glance reveal that the overwhelming majority of these casualties are caused by unilateral accidents. A study by SWOV concluded that 80% of all bicycle accidents consists of unilateral accidents and accidents among cyclists. Moreover, this percentage is increasing. A mere 20% of all accidents involving cyclists concerns motorised vehicles, and numbers are falling steadily. The ADFC article emphasises accidents with motorised vehicles, whereas the main problem in the Netherlands consists of these unilateral accidents, instead of collisions with other vehicles, for example cyclists colliding with bollards.

Dutch cycle-path. Secondary route.
Three metres wide, parallel with
"busy" road. Part of a dense grid
of high quality cycle-routes which
go everywhere.
Anyway, unilateral accidents cause in general only minor injuries. A study by Consument en Veiligheid into unilateral bicycle accidents (November 2008), commissioned by Rijkswaterstaat Dienst Verkeer en Scheepvaart (RWS DVS), revealed that the annual number of treatments at Emergency Departments (SEH) after unilateral bicycle accidents is approximately 46,000. Approximately 6,000 of these victims require hospitalisation. The estimated annual number of traffic fatalities caused by unilateral bicycle accidents is 50. In 2007 there were 189 traffic fatalities among cyclists in the Netherlands.
And finally, improved safety by separation of traffic has by now been conclusively proven in the Netherlands. See for instance the fact sheet ‘Kwetsbare verkeersdeelnemers’ by SWOV (July 2009). This argues in favour of a complete separation of traffic participants with large differences in mass and speed. After all, in serious bilateral conflicts between for instance a car and a cyclist the inequality ratio is huge: 126. This means that in these types of conflicts the number of cyclists injured is 126 times higher than the number of motorists, seeing as people are well-protected in their cars and run relatively little risk at low speeds. In accidents involving cyclists and pedestrians the inequality ratio is a mere 1.7. In accidents involving two cyclists the ratio is of course 1. For the safety of cyclists it is therefore better to separate them from motor vehicles by providing separate bike paths.

Dutch cycle-path. Four metres wide
in quiet residential area
It's also worth pointing out that the view of safety is a bit different here. There are three types of safety to consider if you want to achieve a rate of cycling which is comparable with the Netherlands, or the safety that cyclists enjoy here.

You may also be interested in the concept of Sustainable safety which has lead to a continuing increase in safety in the Netherlands.

I have made a video comparing Dutch and German cycle paths.

Older people cycle in the Netherlands
In addition to the points made by the Fietsberaad statement above, another fact worth pointing out with regard to Dutch cycling crash statistics is that there is a far wider cycling demographic in the Netherlands, including many people who are more prone to injury when cycling and who simply don't cycle in other countries with anything like the same frequency. Retired people cycle for around a quarter of their journeys in the Netherlands and they are the unfortunate victims of 2/3rds of all the injuries and fatalities. Note that the collisions in which these crashes take places almost always involve only the retired person. No other cycles or other vehicles are involved in most cases. Without this more vulnerable group cycling, Dutch injury and death statistics when cycling per km travelled are already the best in the world and they would look about three times better than they do now if it were not for this additional burden of also having the most vulnerable cyclists in the world.

The photos show four of the many different cycle paths within a few hundred metres of our home in Assen, with different types of cyclists using them to cycle with a remarkable degree of safety compared with elsewhere. It's impossible to over-state the importance of a high density grid of extremely high quality infrastructure.


l' homme au velo said...

I commented on the Fietsberaad site and said that Vehicular Cycling is not Safe only the Segregated Cycle Paths like they have in the Netherlands is, and in Ireland we consider it best Practice to have our Infrastructure like theirs.

We the Cycling Community are asking for this but it is a struggle to have this Implemented. Any sort of improvements are hard fought for with massive opposition from the Car Lobby if something is Adopted. Like the new 30KM/H Limit in Dublin City Centre also the Bus Gate and Contaflows for Cyclists.

We have some good People in the City Council who Support Cycling and Pedestrians but we also have Pro Motorist Councillors as well as the AA and the Business Community who are opposed to anything that will slow the Motorists down.

I said that Vehicular Cycling is not Working anywhere it is tried out like Ireland ,the UK,The Us,only something like the Netherlands Works.

Norma said...

Here`s a good exemple from this winter of getting hurt on a seperate bicycle path:
Can you imagine the consequences if this would have happened on an open road? People cycle in all kinds of weather conditions around here and take nasty falls. Fortunately, because of the bike paths, here are no cars around when this happens.

Nick said...

What complete b*ll*cks! There's no way I feel safer cycling alongside cars than I do on the cycle paths round here. Where oh where do these strange people get their ridiculous ideas from?

Neil said...

Part of the problem with the different conclusions (certainly for UK) is that with any separation there will be areas of interaction, junctions, and it is the handling of junctions that is critical to how well a scheme works.

i.e. just taking cyclists away from cars for the straight bits but making a mess of where they cross other vehicles will never make cyclists safer. (it's one of the reasons most experts are against shared use in the UK, because priority is lost).

I can't see how anyone would object to an infrastructure like the Dutch have. On the other hand, I would agree with most objections to paths in the UK. The two are not remotely comparable.

Krakonos said...

Thanks for this informative post. One of the main arguments in Germany against segregated bicycle infrastructure is that cars turning right tend to overlook bikes on these bike paths even though they are painted in red for example. And the statistics here in Germany seem to prove that, I think.
That this is actually a problem in countries like the Netherlands and Denmark too can be seen from the fact that there are very elaborate and costly attempts to address this in these countries too as can be seen here:
It is a bit of an elitist argument though, because many people wouldn't even ride a bike at all, if there is no decent bicycle infrastructure. And that, of course leads to the "safety in numbers" argument.
The overall question is therefore not bicycle infrastructure or not, but can we make bike lanes safe on intersections with moderate effort and is it really necessary to make bicycle lane use mandatory or couldn't we just leave that to the cyclist?
I bet the overwhelming majority would use it even without being required if it is well designed.

Brent said...

I find the Fietsberaad response a bit confusing, but they seem to be including any bicycle crash in their traffic statistics. If a cyclist requires medical treatment after falling, the incident gets counted. Here in Los Angeles, such an incident would probably be classified as a "sports injury."

David Hembrow said...

I'm working on a film comparing German cycle paths with Dutch cycle paths. It will appear on this blog in a couple of days time.

Krakonos: Denmark is not really the place to look to see good junction design. In fact, I've been considering using exactly the post you refer to as an example of bad design in Denmark. Relying on motorists seeing a few LEDs to prevent them from crashing into cyclists as lanes merge is not exactly wonderful. I followed up that post with examples of much better solutions to conflict at traffic lights as done in NL: This one and this one.

Neil: British infrastructure is in the most part really horrible. Worse than the German equivalent. However, British campaigners are a good part of the problem, rarely asking for British planners to even look at the best practice that they could see here in the Netherlands, let alone asking them to copy it.

Anonymous said...

If you closed a road to motorised traffic would it still be a road or would it become a cycle path. I am just thinking it would be the cheapest and quickest way around the problem. Put an end to the iron boxes of death and I will be more than happy to ride on any road.
Cheers, Mark Garrett

Euan said...

There are bike paths and there are bike paths.

Here in Melbourne, Australia, the laughingly named Copenhagen lane on Swanston Street is a disaster. That's only surpassed by the two way separated path on Fitzroy Street which is an absolute menace.

On road bike lanes do not work very well and, overall, cyclists are far better off riding vehicularly. That may change if the infrastructure comes up to snuff, but it certainly isn't the case now.

Anonymous said...

An often-overlooked point is the age of cyclists involved in accidents. I was looking at a PDF brochure by the Dutch Ministry of Transport a while ago that had detailed breakdowns of accidents (part of their yearly traffic safety press releases, I think) but I can't find it now, so I'm reproducing vaguely similar numbers from the SWOV database ( ) even though their latest data is from 2003:

* 53% of all cyclist deaths are age 60 and up
* 17% of all cyclist deaths are children under 18
* 66% of all serious injuries were either under 18 or over 60
* 63% of all accidents resulting in injury or death are those same under 18s and over 60s

In short, accidents involving children and elderly people (groups which traffic planners label as 'vulnerable') far outnumber the number of accidents involving able-bodied adults. This is not necessarily an alarming thing, since these are the groups with the highest cycling rates, and both groups have their own reasons why they're more likely to get into an accident in general. However, these are also the groups that would be most adversely affected by moving cyclists to the main roadway.

I live near a popular cycle route that is used by children on their way to school and elderly people on recreational rides - and I would shudder to think what would happen if cyclists were forced onto the neighbouring 50 km/h road. In fact, I imagine most children would not be allowed to cycle to school, friends or sport clubs, and the elderly couples I see go by on sunny summer afternoons would soon take up knitting instead.

Finally, on a somewhat selfish note, I've gotten used to cycling into town with my hands in my pockets, looking up at the glorious spring sky or looking at offers in the shopfronts gliding by, or maybe having an animated conversation with the person next to me complete with wild, Italian-style gesticulation... on-street cycling would replace that experience with gripping my handlebars tightly and glancing over my shoulder nervously when approaching every intersection.

Michael Radziej said...

I'm also from Germany and a member of the ADFC.

First: the position of the ADFC is not against cycle paths but on making these paths mandatory and banning cyclists from the street. A good cycle path will be used by cyclists voluntarily and does not have to be mandatory.

The problem with cycle paths in Germany is that in most cities money is a scarce resource. Politicians want to be considered bike-friendly, so they encourage cycle paths, but only spend the money absolutely necessary to build anything that can be presented as a cycle path. And often the motivation is more to get bikes from the street so that car drivers can go fast.

Within cities, cycle paths are often in very bad shape, narrow, bend in strange ways, not separated from pedestrians, and not properly safe in crossings. You often have problems with pedestrians and with turning cars, or cars backing out of their garage. You need a lot more concentration if you use them, compared to using the road, and you usually have to slow down quite a bit. You often cannot overtake other cyclists.

Out of cities, cycle paths are mostly only on one side of the road without proper guidance for cyclists that have to cross the street to use them, and on crossings the way of right is routinely given to the cars, even turning ones. In crossings with a lot of car traffic, it's usually better to dismount and walk your bike across, since then you have at least right of way regarding turning cars. Sometimes right in the middle of between two villages the cycle path changes the side of the road.

In some cases, cycle paths just have a dirt surface that changes into a muddy puddle after a rain, and you cannot legally use the road. Some times you have to share a path 1,20 m wide with lots of pedestrians, especially on weekends. And you cannot legally use the road.

So the bottom end is: It would be fine to have an infrastructure like the ones pictured on this blog. As long as we have not, I want to have the right to ride alongside the cars. It is safer than using bad and narrow cycle paths, where you are out of drivers' views and get into danger at each and every crossing.

That's the background.

Anonymous said...

just make this: "cars turning right tend to overlook bikes" a capital offence of you do it more than once, and if you hurt the bicyclists. Then pure darwinisme should remove those blind car drivers that can not see.

Anonymous said...

Yeah, those blinking lights in Copenhagen is not the right solution.

But how do you make safe junction when you have a major road junction? the big road here has 3 lanes in each direction, it has 1 right turn lane and 2 left turn lanes. The other smaller road has 2 lanes in each direction, + 1 right turning lane. It is one of Copenhagens big junctions.

And no, you can not really make people use another path to travel. You see, this junction is close to a bridge across the habour, over to one of the islands that makes up Copenhagen. I bike past here every day to get over that bridge. There are 1 other shared bridge not to far from this, but besides being a detour, i have to turn more often to take that path. There is a bicycle/pedestrian bridge in the opposite direction, but same distance. But to get to that i have to drive along the habour, or go down some stairs.

There is a 3. shared bridge further kilometers away, and even 1 freeway bridge 10 kilometers away.

i dont think that Denmark are ready to put up with the dutch solution of bicyclists having their own light. Besides, since most go straight here, and there are not so many bikes from the sides, do they really need it?

I would like to see this road digged down under the city center, but i dont think that will happen either. We could make a bicycle bridge to take you over this junction. Or under, but will that make people feel safer?

Anonymous said...

that video shows why you need 3 wheels in the winter

David Hembrow said...

For the benefit of anyone following these comments, please see the video comparing Dutch and German cycle paths.

Severin said...

A lot of people in the US, or at least LA, are under the impression the vehicular cycling is the way of the future and all safety issues will be addressed by slapping sharrows in every street.

Anonymous said...

The mind boggles... What braingymnastics must these people have performed to come to the 'conclusion' that bicyclepaths are less safe than vehicular biking?! And those statistics! I can't wrap my head around those. I mean, in this little country of ours, more people cycle more miles in any given year than Britain or Germany, or even the USA. Suppose during one year three Dutch cyclists die (opposed to thirty German, British or American cyclists that year), and two of those died due to a freak accident with a car (it does happen, if rarely) does that mean that according to statistics a whopping 66 percent of Dutch bicycling deaths occurred as a result of a collission with a car? It gets even worse if that year had one extra death-by-car. That would ratch the percentage up to 75 percent! Wow! Those unsafe bicylce paths! Better throw cyclists in front of sharks, uh, cars in true 'road warrior' style. Of course the actual *numbers* of deaths will multiply, but the *percentages* will go down!

Now, how did that quote go? Something about there being lies, damned lies and statistics?


Anonymous said...

David - living in the UK (Glasgow), and using my bike as much as I can in a everyday, normal clothes and no helmet manner, I can only weep when I see the differences between cycle-friendly countries and here. Cycle lanes here are a joke and road traffic is too fast, especially in town/residential areas. Many cyclists "adapt" by becoming as fast and aggressive as cars, wearing lycra and all sort of specialized gear, making cycling more a hobby/sport than a easy way of transport. the authorities don't miss any chance to exort cyclist to wear helmets, reflective jackets and generally blaming them in case ot accident for "not being visible enough" - a very tragic case here is Scotland only a few days ago.
Feeling very disillusioned (but will keep cycling my way!)

tOM Trottier said...

The stats on the Fietsberaad show nothing useful at all. I commented there that what should be compared is the number of fatalities or hospitalisations, not "accidents".

As for vehicular cycling, in Canada, to get anywhere/everywhere, you need to be a vehicular cyclist. While it is scary for most, and there are some serious accidents, if the speed differential is small(10-15kph), the drivers have the time and space to treat cyclists safely.

The most common thing a motorist says after an accident is that she didn't see the cyclist. If you place yourself in a visible position and wear bright garb, you avoid most of the problems.

Courtesy and defensive riding are also important tools.


Taliesin said...

I realise that is rather late to comment, but was prompted by yet another cycle path bashing session on the CTC forum. Basically, VC proponents will claim that the Dutch are worse of than they would otherwise be because of the presence of segregated lanes which they are convinced bring a 3-11X greater risk at junctions vs on road cycling. (

So I guess the question is, looking only at KSI bicycle accidents involving motor vehicles, do the accidents cluster around intersections with segregation? I have a suspicion they may, simply because it is easier for a KSI to occur in the presence of faster traffic moving through an intersection. The consequences of an error or stupid act by anyone are likely to be greater, than on a traffic calmed intersetion. Is this why we cannot be simply told that the Dutch are safer on their paths than on their roads, at least in terms of KSI stats?

Euan said...

@Taliein: Intersection collisions are part of the problem, don't forget that 80% of bicycle accidents don't involve a third party; car, pedestrian, cyclist or otherwise.

In Holland and Denmark it is illegal to have an against flow bike lane, so the worst case scenario is 4x, not 11. There are no figures in the Wiki article for Holland, but really it doesn't particularly matter.

Fact is, people like bike lanes. I don't. I find them inferior to the roads in most cases I'm a cycling enthusiast. Ordinary people like bike lanes and it's ordinary people we want riding, because the fairy insignificant increase in risk with bike path/road collisions is outweighed by the safety in numbers effect and the increased health of the people who do ride every day.

It's reckoned that a 20 minute ride at a bike path pace a day is enough to keep one healthy. I don't know the truth in that but the dutch have a much lower health bill than the UK, as a population they're less overweight. While there may be other factors than cycling at play evidence from the USA suggests that a state's obesity has a correlation with the amount of cycling that happens.

As I said earlier, I don't like bike paths for my own cycling but the evidence is hard to ignore; bike paths get people riding.

David Hembrow said...

Taliesin: It's rather difficult to make a direct comparison, as you'd have to look for large groups of cyclists disobeying the law and riding on the road. You don't find that here, as it is extremely rare that riding your bike on a road vs. on a cycle path would actually bring you any advantage at all.

The comparisons made in an attempt to discredit cycle paths are on hypothetical "facilities" which don't correspond to real life - at least here in the Netherlands. There are differences in the ways that junctions are designed which are ignored by that rather biased article.

Junctions here are designed to reduce clashes and prioritize cycling. You can see a few examples here and here. Note in particular how cyclists are separated in both space and time here, here, here and here. These types of junctions, the norm here, are completely ignored by the article.

Euan: I've yet to meet anyone who is anti cycle paths in the Netherlands. It would make no sense. You can make faster and more convenient journeys by using the cycle paths, and I suspect you would do that too if you were here.

I do agree, of course, that low quality provision is often less good than none at all, at least for existing cyclists who are brave enough to ride on the road.

However, I think it's a mistake to see cycling infrastructure as only a crutch for the less able. If done well, then it improves conditions for everyone.

The differences between Dutch cycle paths and those built in other >countries, are not small differences. A growing network of intercity superhighways are used by many long distance commuters, and cycle paths are of such quality that you can have competitions on them, for example a time-trial in which the fastest competitor averaged 59 km/h over 51 km. Is that quick enough ?

Dutch infrastructure made a huge difference to my commuting speed.

Aurora said...

Come on, what's wrong with accidents? I can break my own nose by falling down in my garden, so when I fall off a bike I can break my arm. ANd I actually like seperated roads, because otherwise I would have to ride my bike next to cars who still drive well.. let's say.. min. 200% faster than I do. That's no fun, really it isn't. Still, I better like to actually bruise my (i don't know) my ankle than have to go by bus.