Tuesday 16 September 2008

Three types of safety

Separate from the main road, a family cycles together, side-by-side. Straps on the front child seat are not done up. Slightly older children ride their own bikes. No-one wears safety equipment. The child in the rear is fiddling with something but the parents have no concerns concerned. This illustrates Subjective Safety.
The Netherlands is the safest place in the world to cycle. This is sometimes put down to a "safety in numbers" effect, but actually the infrastructure design is the vital component and shouldn't be overlooked. A lot of people would like their own country to emulate the Dutch success, but often they don't realise what is needed.

I used to do cycle promotion work in the UK, travelling from city to city and talking to a great number of people about cycling. They all already knew that cycling was healthy, good for the environment etc. Many people would like to be able to cycle. The number one reason that the average person in the street would give for not cycling was "it's too dangerous". So, what did they mean by this ?

One of many four metre wide cycle-paths
in Assen. School run cycling.
There are three measures of safety, all of which have their place in Dutch bicycle provision:
  1. Actual safety - How many km you can expect to travel before you're injured on your bike.
  2. Subjective safety (sometimes called "perceived" safety) - Are you near fast moving traffic ? Is it easy to make a turn across traffic ? Do you have to cycle "fast" in order to keep up ?
  3. Social safety - Is there a mugger around that blind corner ? Will I be attacked in the street if I cycle ?
Cycle campaigners and planners might interest themselves in the actual safety, and it's a good thing that they do. Cyclists should of course be as safe as possible. However, no-one really makes a decision on whether to cycle or not based on these figures. Actually, cycling isn't really very risky in most countries and these figures often feature in cycle promotion literature. However, they're not successfully convincing people to take up cycling.

When people make the decision about whether it is "safe to cycle", they generally mean the second and third of our three different types of safety: Subjective Safety and Social Safety.

Also, if they're making a decision for someone else - perhaps their child or their partner - these issues become even more important.
Mother and children. Having several
metres of separation from the road
is essential for a high level of
subjective safety.
How do you improve Subjective Safety ? Here's a partial list:
  • Cyclists should never mix with high speed or high volume motor traffic. A third of all roads in the Netherlands have a speed limit of 30 km/h or lower, most 50 km/h (30 mph) roads provide cyclists with a segregated path, as do many 30 km/h roads with higher volumes of traffic.
  • On-road bike lanes and cycle paths without sufficient separation from the road are not suitable with high speed or high volume motor traffic.
  • Reductions in speed and volume of traffic always help. All residential streets and a third of the entire road network has a 30 km/h (18 mph) speed limit or lower.
  • Fully segregated cycle paths provide a good degree of subjective safety but must be built to a suitable standard. In this area they have a minimum width of 2.5 metres if for single direction use and 4 m for bidirectional use. Paths for pedestrians are separate.
  • Junctions should be designed to make sure that cyclists are not left out. Cyclists can be separated in both time and space from motorists to maximise both efficiency and safety.
  • In many cases, cyclists avoid junctions altogether so they cause as little inconvenience as possible.
  • In Assen, the new standards require that cycle paths which follow the line of roads are separated from them by 2.5 metres. Where this isn't possible you will find a metal barrier is used, to provide a feeling of subjective safety as well as actual safety from crashing vehicles.
  • Where possible, cycle routes follow a completely different route to driving routes, which of course improves the feeling of safety further.
  • Reducing the noise of motor vehicles by using quieter road surfaces and installing noise barriers between the road and cyclists helps.
In the countryside away from all
motorized traffic, our family, rides
in comfort,
For social safety:
  • You should always be able to see out of any tunnel as you enter it.
  • Blind corners on paths are not acceptable.
  • Cycle paths should be wide to allow cyclists to move out of the way of other people who may be on the path.
  • Good sight-lines are required and there should be no places to hide along a cycle route.
  • A low crime rate and a good conviction rate are needed. Cyclists should not feel that the police do not take their complaints seriously. 
  • Areas that are clean, litter free, graffiti free, where grass is mowed and plants are not allowed to overhang the cycle path have a better feeling of social safety.
  • Cycle paths should be lit at night so that you can see potential muggers, obstacles on the path etc.
If subjective and social safety are improved then people will cycle. They will want to. and so they will do it.

To summarise... No-one will do anything that feels too dangerous to them. Everyone wants their child to be safe and their partner to be safe. That's why so many journeys which ought to be cycleable are made by car. There is no point in arguing with people's decisions, or ridiculing them. The person making the decision to use a car has made it for quite logical reasons. Their level of confidence about cycling in the conditions around you is not the same as your own.

What to do... If you want people who do not cycle to take up cycling, then the right thing to do is to campaign for or design in road conditions which make cycling into an appealing option. That is what the Dutch have done. Everywhere. It is the key to the high cycle usage and high cycle safety figures.

Don't make the mistake of thinking that subjective safety is a concern only for inexperienced cyclists. No-one suffers from cycling being pleasant. Steps to increase the subjective and social safety of cyclists lead to a better cycling experience for all. Experienced cyclists are less likely to give up cycling in a subjectively safe environment. It becomes a lifetime habit. People continuing to want to cycle on the road when there is a parallel cycle-path are a sign of failure due to low quality. Always set your sights high enough when campaigning or planning.

A British child being trained. Wearing
fluorescent clothes and a helmet to ride
in the gutter on a residential street
which is subject to rat-running
So, where do helmets and fluorescent clothing fit in ? For some individuals, wearing such a thing improves their own feeling of safety to the level that they will ride. However, these items actually do little to improve actual safety and can have a negative effect on the subjective safety of other people due to making cycling look dangerous. Where cycling has a high degree of subjective safety, as it does here, no-one wears these safety aids. Dutch cyclists are safer without them than cyclists elsewhere are with them.

For more on the same theme, perhaps this post is most suitable. There are quite a few other posts tagged subjective safety which show different aspects of what makes cycling subjectively safe and the result of it. Amongst the other things needed to make cycling attractive is to make cycling more direct, so there are a lot of examples there to illustrate that concept too.

Big picture vs. small picture subjective safety
December 2011 update: I've realised since writing the text above that many people have slightly misinterpreted what I meant. There are two types of "subjective safety".

The most important, "big picture", type of subjective safety is that of society as a whole which causes everyone to cycle. This is what my article is about and this is what causes Dutch people of all ages and social positions to find cycling to be a safe activity, even wearing black clothes on an unlit bicycle at night. Cycling in the Netherlands always feels safe. This leads to a very high modal share for cycling.

The other, "small picture", type of subjective safety is that which involves people using safety equipment such as fluorescent clothes and helmets when they cycle. While this may make it possible for a few individuals who are already interested in cycling to cycle a bit more and to feel safer as they do so, it is debatable whether this does anything for the overall modal share for cycling.

Some people argue that it is detrimental to the modal share if cyclists take up visible safety equipment to improve their own "small picture" subjective safety. I suspect that "big picture" subjective safety is barely altered at all by this. The issues that people who don't cycle can see and which put them off cycling are with street design, not with the clothing of existing cyclists.

All the photos are of subjectively safe cyclists in the Netherlands, except the last which is of a British child being taught "safer cycling" in Cambridge. He is riding too close to the kerb on an unsuitable bike, but wearing fluorescent clothing and a helmet. The child in the fluorescent clothing will probably give up on cycling within a few years, while the Dutch cyclists will in all likelihood keep on cycling through their lives because high levels of subjective safety make it a pleasant thing to do.

The children in the photo at the top are all tired. Why ? Because this family is returning from the last day of a four day cycling event. They've cycled at least 40 km per day as part of the event, plus however far it is between their home and the startplace in Assen. The object which the child at the back is fiddling with as he cycles is a medal.

Per km travelled, Dutch cyclists are 3x less likely to die and 4x less likely to be injured than those in the UK, 5x less likely to be killed and 30x less likely to be injured than in the USA (statistics found here, page 506). Read more about the safety of cyclists in the Netherlands. However, this doesn't tell nearly the whole story, it's actually much safer than that for Dutch cyclists who fit the usual demographic for cycling in other countries. Two thirds of all cyclist deaths in the Netherlands are of people aged over 65. Most of their lethal injuries come from single vehicle collisions. i.e. when old people fall off their bikes, old age sadly can be enough to do the rest. This effect is virtually unknown in the UK and USA because so few older people cycle.


cocosolis said...

Thanks for this post - I find it the clearest analysis I've come across so far as to why cycling is 'normal' in other countries, while it remains marginal in the UK. British politicians and planning officials meanwhile delude themselves by referring to patches of white and green paint on the road as 'segregated' cycle lanes - they are nothing of the sort! And yet they quote the number of miles they've painted as evidence that they have 'tackled' the problem. "The present government has done more to encourage an increase in cycling, through a wide range of policies, than any British Government in living memory" (David Chaytor, my local MP - see: http://manchestreker.blogspot.com/2007/05/david-chaytors-response.html)

dale said...

My city has a population of 400K, about 1300/km2, yet only 150km of separate bike paths.

Yes, having bike friendly road construction codes are needed to improve the subjective safety factor in the future. But living in the present, one must ride on roads and neighborhood streets to commute to work, stores, etc. With our poor biking infrastructure, route selection is key, coupled with education on how to ride safely when cars are present.

I disagree with the nonuse of helmets. I have broken two helmets from falls while riding on roads and bike paths over 6 years. Helmets are a wise choice to limit traumatic brain injuries.

David Hembrow said...


I fear you've slightly misunderstood what the article is about. I'll summarise here: Where cycling has been made to feel safe, "normal" people will cycle in large numbers. They also no longer feel the need to take what then becomes an irrational step of dressing up in "safety" gear to do something that is "safe".

We have had many visitors who wear helmets when cycling in their own country who remove them when they cycle here. No-one tells them to do so, they just no longer feel they need them. If you try it for yourself, you may well find you feel the same.

I've looked at the omaha trails on the web and see that their emphasis is very much on recreational use. A quote from the website says "you probably have a trail less than ten minutes away from your own home!". This is not at all like the approach of the the Dutch who provide a finely spaced network of cycle paths).

We've got recreational trails too, and they are very pleasant for "going for a ride", but they're not really part of the practical network in the city here either.

I'm not against you wearing a helmet if you feel happier with one. If it reduces your perceived risk such that you ride your bike, then that's good. However, please recognise why you wear it.

It's got very little to do with cycling being dangerous in and of itself.

There is no reason why cycling needs to be any more unsafe than a lot of the other activities that you and I take part in every day without a helmet. These include walking, climbing stairs, riding in automobiles...

The justification for a helmet for bicycle use only is one of compensating for the risk (or perceived risk) around you due to your local environment, not of compensating for any inherent risk of cycling.

While cycling is genuinely much more dangerous in the US than it is over here, that is due to other factors, such as car oriented design of streets and driver behaviour, not due to any inherent danger of riding on two wheels.

Sport cyclists, including MTBers, road racers, BMX riders, do wear helmets here at least some of the time. They take risks, and they're compensating for their additional risk. However, sport cycling has little to do with utility cycling.

inconvenient_truth said...

Excellent post, David. A clear summary of the way people (mis)understand "safety".

As someone living in two communities, one German and "cycling friendly", and one British and "cycling wannabe", I see day to day these misunderstandings about cycling safety. They are largely based on experience being limited to "the way it's done here".

Sarah said...

This is really interesting.

Subjective safety is so important. I recently started cycling more at night around here (Boulder, CO, U.S.A.) but my subjective safety is only moderate - bike paths are not lit, one of the places I go regularly involves a neighborhood with no street lights at all, and there are few other people out cycling on the same routes. I feel the need for bright lights and reflective clothing (not to mention a helmet) to ensure the cars I have to share the road with see me.

Regarding helmets: I was just looking up statistics about cyclist deaths and helmets in the U.S. and they make helmet wearing sound like a very rational choice given the way cyclists have to share the road with cars in this country. However, it sometimes starts to sound like helmets are proposed as THE solution to motor-cycle accidents, instead of people looking further and realizing that a better solution is to redesign the routes the cyclists take! It's as if there is an acknowledgment that bikes sharing the road with cars is dangerous, but there is no attempt to solve the real problem, just to slap a band-aid on it and say, therefore you should wear a helmet. So I think I get what you mean about helmet use. I wonder if I would feel safe enough to take my helmet off in The Netherlands, or if it is too ingrained in me...

One more thought about bike path design (sorry for the long comment!). In Boulder there are bike paths, including one along a major road. This seems like a good thing; I think it does encourage more cycling, but some people perceive it as more dangerous than being in a bike lane on the street (there are no bike lanes on this particular street, however). I think the reason is that the path is not well-designed when it comes to intersections with roads. The cyclists have to act like pedestrians at these intersections, and depending on which direction they are going, may be against the flow of traffic. Cars turning right or unprotected left (such that they need to cross the path) do not always expect cyclists - and thus the onus is on the cyclist to slow down and verify that the car has seen them. We could do well to emulate The Netherlands, that is for sure!

Anonymous said...

Excellent post. I fully agree that cycle promoters here in the UK pour far too much focus on purely statistical safety.

To throw another helmet into the fire of the safety debate: after many of going pretty much everywhere by bike (started at uni and never saw a good reason to stop) I finally picked up the eye-opening book City Cycling by Richard Ballatine. Among other issues, he highlights ways to cycle more safely, to control the vehicle behind you, and generally control your risk. I feel even safer on the roads now I'm implementing tips from this book.

David Hembrow said...

Thanks Adam. Unfortunately, for all their "focus", the UK's roads are not particularly safe for cyclists. What's more, it's worse now than it was 50 years ago.

NL on the other hand not only feels safe, but is actually also the safest place in the world for cycling per km travelled. The infrastructure here was redesigned in order to bring about this increase in safety, and it has worked.

The problem with all these ideas about controlling vehicles behind you is that they don't work 100% of the time. Some people will run into you on purpose on Britain's roads. I've experienced it myself.

Jayadeep(JDP) said...

David - thank you for this informative post. I am a new commuter(2+2KM only) in Bangalore, India. The safety perspective is very different from the way things are in our country. My thoughts are at http://tinyurl.com/b24unm. Feel free to comment on this.

David Hembrow said...

Jayadeep, thanks for your message.

I have no experience of Indian conditions, and they're obviously somewhat different to the European experience that I have had.

There is a difference between subjective safety as a rule for all, and maximising your own feeling of safety by your own behaviour. For many years I did just as you do on the unfriendly roads of the UK, but this did nothing for the cycling rate in the UK. The majority of people in the UK still never cycle at all.

The approach here in the Netherlands is, as you've noted, somewhat different. Cycling isn't just for those brave or desperate enough to cycle, but for absolutely everyone, of every age, race and both sexes.

It's that sense of subjective safety for all that I am most interested in, rather than simply making cycling OK for oneself.

It's a matter of whether you would encourage your children or elderly relatives to cycle, or whether it seems to be something which is suitable only for fit young adults.

BG said...

This is an old post, I know, but I'm really interested in your perspective here. I agree with you almost all the way through -- indeed, perhaps all the way through, but there's one thing that has me curious. Toward the end you say that "No-one suffers from cycling being pleasant. Steps to increase the subjective and social safety of cyclists lead to a better cycling experience for all. Experienced cyclists are less likely to give up cycling in a subjectively safe environment." So, what exactly does happen to the person who wants to ride like an American "vehicular cyclist" on the roads (not paths) in the Netherlands? I ask this as a genuine, not rhetorical, question. Do car drivers there understand what the cyclist is doing if, say, he takes the lane in a narrow stretch and prevents cars from passing? If I want to ride fast (in my case, not because I'm a racer -- I've never raced in my life -- but because I tend to get up late for work), can I do it? Or am I stuck on the bike path with the kids and grannies? I assume that you ride your velomobile on the road, not the path; how do drivers react? Here in the US, many people oppose Dutch-style separated facilities because they think they will give the pro-auto forces the excuse they've always wanted to ban bikes from the roads. How have the Netherlands avoided that reaction?

David Hembrow said...

BG: The question you ask virtually never arises because there is almost never a reason to want to ride on the roads instead of the cycle paths. The paths provide direct and fast routes for cyclists. The roads provide more traffic lights, traffic calming etc. The roads provide altogether a less convenient way of getting about.

I commute, using either my two wheel recumbent or a velomobile, on the cycle paths. By doing so I have priority where the roads and cycle paths meet, I get a shorter route than I would by taking the road, and I can ride flat out for almost the entire distance if I wish to.

There are examples on this blog of how the cycle paths are made so convenient to use for cyclists.

Having said all that, occasionally I've found myself on the road due generally to having made a mistake. I've never had much bother from Dutch drivers.

tedsfiles said...

Great and interesting article. I have to agree that most here don't cycle due to perceived risk. Unfortunately for Sydney, it's an actual risk also, since the cars here are a.holes, and we have very few paths, so must mix with angry cars.
I have to disagree with the statement about fluoro vests not improving safety. Here one of the main modes of accidents is being rear ended by a car. I always check out other cyclists to see how visible they are. When you're riding solo, and there's no other bikes about, dark clothing can make you quite invisible. So fluro helps a lot. It also makes riding seem more popular, since those riders are so noticeable.
I'd love to have infrastructure where helmets and fluoro has no effect, but we're a long way off here in the city that hates bikes.
You'll be pleased to know that they're currently constructing 200km of paths in the middle of the city, so fingers crossed.

Anonymous said...

Good post, as usual.

Regarding helmets - I used to be in agreement that helmets aren't necessarily safer, but are merely perceived as such. Until I had an incident where my helmet saved my life.

It won't help much if you're hit by a motor vehicle travelling at speed. But it makes a huge difference in the much more common circumstance of a wipeout.

I was doing something silly - trying to balance at low speed, on a narrow bit, when I hit a leaf which, I suspect, had some ice underneath it. I'd worn my helmet out of habit, as I'd actually planned to ride without it. I went down so fast I couldn't brace myself.

My torso hit my handlebars as I fell, and whipped my head onto the concrete. I hit squarely on my temple, with a loud crack that sounded like a gunshot. I did some serious damage to my knee and arm, and I suspect I gave myself a minor concussion. But I'm absolutely sure I'd either be dead or have suffered serious brain damage if not for my helmet.

Thanks, Bell helmet. No you, no me.

David Hembrow said...

Jim: I have no desire to stop you wearing a helmet if you like to do so. However, your statement that your are "absolutely sure I'd either be dead or have suffered serious brain damage" is simply illogical. Neither you, me, nor any doctor can say this with any certainty.

There is only one way to know whether in your particular case a helmet really would have made such a difference, and that is to set up an experiment with your hypothetical twin brother having exactly the same crash as you did, but without a helmet, and seeing whether he was any worse off than yourself. That's how science works, not by guessing based on a feeling that you think it helped.

People fall over in all sorts of odd ways all the time, not necessarily when cycling, often when they are not wearing helmets, and almost always without causing brain damage.

I don't often post about helmets, but I did make a post specifically about the risks to dutch cyclists due to not wearing helmets.

Charles Martin said...

David, The main theme of your campaign to export the Netherlands cycling experience to the UK is that the provision of cycling infrastructure will encourage greatly more trips to be made by bicycle. Yet this does not seem to be the case in the UK even when excellent cycling facilities are provided. Between the ages of 4 and 37 I lived in Stevenage, and was a keen cyclist throughout those years. However, I had the fully segregated cycle lanes virtually to myself. The 2008 Stevenage Transport Strategy, quoting figures from 1991, states that 4% of residents were cycling to work. This is obviously better than the UK average of 1%, but nowhere near Netherlands levels. I fear that there is a deep-rooted aversion to cycling in the UK that even near-perfect infrastructure does not overcome.

David Hembrow said...

Charles, I'm going to cover Stevenage, Milton Keynes and similar places when I get a chance. An attempt was made in these places, but in my view it's not a good enough attempt. There are several ways in which it differs from what is normal in the Netherlands.

StefanLangeveld (Amsterdam , The Netherlands) said...

@ BG (Or am I stuck on the bike path with the kids and grannies?):
In NL cities, all cycles including scooters and the wide children-movers must try to ride on the cycle paths, whilst avoiding cycles from opposite direction. Cycling on the car lane is difficult because cars must slow down to pass a bike; nobody does, i rarely do it (but want to).
Anyone saying that Amsterdam is the cyclist captital is blind / toeing the pc line / working as a cycling consultant touring the world exporting this error / local politician.
In recent decades, cycling has become far less pleasant for all. The paths are not the only problem. The city is full of obstacles for cyclists: posts, humps, impossible corners.
Only recently people are waking up, at least they complain about the scooters on the paths (who were forgotten by the cycle path proponents?). It will take longer before they realise all such problems are caused by cutting up and allocating the space.
David H. has experience in smaller towns where the paths are less crowded.
More on my Baluw.nl (Items in English : A warning against Dutch cycling experts)

David Hembrow said...

Stefan: It's not perfect here in the Netherlands, it's merely better than everywhere else. If you actually knew how it was to live and to cycle elsewhere you'd have a better appreciation of what you've got.

Who is "working as a cycling consultant touring the world" ?

Colibri said...

So there does exist wanna-be "vehicular cyclists" in NL ! :-)

Stefan: sorry but I'm no big fan of the exemple on your website about Morocco.
Maybe because I don't like risking my life just to be able to cross a street, among steel monsters...

PS: road casualties amount to 4000 per year in Morocco.

Unknown said...

Thanks so much for drawing such a concise picture of what the real problems are with the cycling infrastructure here in the UK. Having been a cyclist for many years (and having been hit by a car on the street in London), I can attest to what is a legitimate "fear for safety". Discussing the issues on a blog page is great for sharing informatin and ideas in the format of preaching to the choir, but how do we make these needed changes a reality? It seems to me, without a united approach from cycling campaigners (including support from other stakeholders) the right solutions will continue to fall on deaf ears. You have the best suggestions I've seen so far.

Stephan Matthiesen said...

Thanks for this post - it's very useful and important to discuss social safety; actual safety, and to some degree perceived safety, are already talked about a lot. I know a few places in Edinburgh where it's really all about social safety, e.g. paths through derelict industrial areas or some housing estates. Nothing ever happens there so they are perfectly safe, but your mind's alarm system is totally on edge because of all the factors you mention, so you avoid them and take a route along a busy road that is *actually* more dangerous.

I would like to add one feature to your list of infrastructure that impacts social safety: Chicanes and other barriers. Here in UK they are everywhere.

They are a disaster for social safety because your mind feels that they block your escape routes: if somebody tries to mug you, you can't cycle away quickly. Chicanes are also objects that teenagers like to sit on; most of them of course perfectly nice and harmless people, but seeing unknown people sitting on a barrier across your path still triggers your fundamental human alarm systems.

Some chicanes are of course impossible to negotiate or even in a dangerous state, but I personally find even good, wide chicanes somewhat alarming in some areas, because I feel trapped by them.

Unknown said...

An excellent post. I agree entirely with your summary of the 3 types safety. I work as a health & safety supervisor for an oil company, and the concept of subjective safety has never been raised in corporate literature or procedures - I now think it ought to be.

I live in SW England but have just returned from a bicycle tour through parts of Belgium, the Netherlands and NW Germany, and I have seen exactly what you are talking about in the Netherlands; the difference in subjective safety is so obvious once you have experienced it.

I have seen a lot of posts on various cycling websites & blogs quoting examples where "if I hadn't been wearing a helmet I would have suffered injury", etc. However, in most of these cases the root cause of the accident appears to have been the fact that the cyclist was cycling on the road in the first place, mixing it with motor traffic and having to deal with road furniture and layout designed for motor vehicles. I didn't wear a helmet on my tour and never felt the need for one.

Multiparty Democracy Today said...

Out of all the statements about what is an acceptable number of traffic major injuries and fatalities, I think this one resonates with me the most. That number, is 0. It's 2015, how much longer do we have to wait? https://www.youtube.com/watch?v=bsyvrkEjoXI&index=1&list=PLl2p5jFnfRhF4nt7O96lfAKW8EJartJeN. I am guessing that about 70 people died in Melbourne in the year this video came out, and to think, that is multiple classrooms full of people. That is more than a city bus or two. That is what that video proved. And they included several children, to represent those who died. And especially if they were in the passenger seat of a vehicle, were being carried by their parent, or on a seat of a bicycle, they had no fault in their death, yet we blame the victim or car driver for every crash despite the fact that almost every collision in 1972 that would otherwise occurred has been reduced in some way by infrastructure of the roads themselves. Cities who do not adopt Vision 0 despite the obvious fact that it is an attainable goal, you sicken me.

Dwayne said...

Fully separated cycle-paths may be the key. Our city, Houston, has recently implemented and invested in hike and bike trails and the difference is huge. I have my son in a trailer behind me and when we go for a ride I get on to the hike and bike trials almost immediately. Large numbers of walkers are also using the trails and I believe that it has created a type of culture around the trail system. I look forward to more of these types of trails coming to our city. Riding in them is fun and faster then sidewalks. Streets dont seems safe being shared with cars.