The myth of the "tipping point" and the fragility of cycling.

Birmingham once had much bicycle production
and use. However, like the rest of the UK
this city's cycling declined from the 1950s

It has become popular to make statements about cycling somehow taking on a life of its own and growing without further investment once a particular modal share has been reached. A fairly recent example of this sort of thinking appeared in a grant application document from Birmingham City Council:

"Birmingham is working towards the ‘tipping point’, a common pattern within cities, where a modest rise in cycling levels suddenly gathers pace. We want to accelerate the pace of growth further, creating a visible ‘step-change’ in levels of cycling within our city being part of everyday life and mass participation a reality. Our aim is to achieve a cycling modal share of all journeys of at least 5% by 2023, which research undertaken by the European Platform on Mobility Management (EPOMM) has shown is sufficient to generate the critical mass required to make it an attractive mode of travel. By 2033 we want this to rise to levels of comparable European cities such as Munich and Copenhagen at over 10%."

As is so often the case, they're aiming far too low. A target of just 5% of trips at a point in time ten years in the future ? Attempting to achieve such a slow rate of improvement makes it difficult to measure whether there has been any success at all year on year. It's also a good way of ending up making no progress at all. Nevertheless, this is described as a "step-change".

It is also odd that their aim over 20 years is to emulate countries which have achieved less than the Netherlands, and also that they define "comparable" with Munich and Copenhagen as a cycling rate of 10% of journeys when both those cities are currently at roughly double that level.

But the biggest error is the reliance on a "tipping point". Where is the evidence for the existence of this "tipping point" ? Actually, history shows us that without continuous substantial investment to support it, cycling declines even from a very high modal share.

Examples of decline from a high level

Before 1962, the British made more journeys by bicycle than the Dutch do
now (as a proportion of all distance travelled)

In the Netherlands, 27% of journeys are currently made by bicycle. Because it is mostly shorter journeys that are cycled, that translates to bicycles being used for around 10% of the total distance travelled.

In the late 1940s and early 1950s, British people used bicycles for around a third of the total distance that they travelled. This declined steadily and it took until 1962 before the UK dropped to the present day Dutch level with about 10% of distance travelled being by bicycle. Since then, the UK has declined and stagnated.

What happened in the UK to make people switch away from bicycles was a huge programme of investment in infrastructure for motor vehicles while bicycles were mostly not catered for at all. Cycling became less safe compared with other modes of transport as well as subjectively unsafe due to the proximity of an overwhelming number of motor vehicles. Cycling became less desirable as a mode of transport and has become marginalized.

New Towns in the UK provide an example in miniature. Stevenage, for instance, had a higher cycling modal share in the past than it does now. When it was first built, there was (for the UK) a relatively good grid of cycle facilities and these led to most locations. Decades of under-investment, lack of maintenance and not bothering to integrate cycling into newer parts of the town have resulted in cycling having no advantage in Stevenage and that town now having roughly an average cycling modal share for the UK.

The top line shows cycling in the Netherlands by year
The second line shows Denmark. Cycling has been
in decline in Denmark for 20 years

Denmark provides another example. In the 1980s, Denmark had a cycling modal share which was slightly behind, but similar to that of the Netherlands. Both countries were investing similarly and growing cycling at a similar rate.

It was believed that the Danish culture would result in them always cycling. In fact, cycling comes through investment in making cycling into the most attractive option.

Unfortunately, because investment in and prioritization in planning for cycling were not maintained at an adequate rate and the result has been a steady twenty year decline in cycling in Denmark.

Davis in California, which calls itself the "Most bicycle friendly town in the world", is a small city with a size population to Assen (though it's much more densely populated than Assen). The top cycling city in the USA, Davis hosts a large university for its size as well as other educational facilities. A high student population always make it easier to achieve a high rate of cycling and Davis has a high student population even compared with other university cities. While one quarter of Groningen's population are students, and the population of Cambridge in the UK consists of one third students, more than a third of Davis' population are students and a large proportion of the rest of the population are associated with education.

Davis once described itself as "home to 15000 bicycles", but that was when the population of the city was smaller and even more focused around the university. Some people estimate that as many as a quarter of all trips were by bike in Davis in the 1950s but there has since been a well documented decline in cycling to the point where the cycling modal share is somewhere between 1/4 and 1/3 of what it was, with even students cycling less than they once did.

You may wonder how this could have happened. An interesting reply to the last link about Davis points out that they have experienced demographic change which works against cycling (fewer students as a percentage of the total) and many other changes including the retirement of a key city figure with the result that "the city lacks any upper level administrators who are anywhere near as dedicated" to cycling.

Davis is now trying to boost cycling by restricting student car ownership in a similar way to Cambridge - something which cannot be applied generally to all cities, and which won't help much when recent growth to the city's population has largely been due to non-students.

More recently, there's the much documented fall of cycling in China. The term "critical mass" was coined by the film-maker Ted White after he saw Chinese cyclists form into a mass at the side of the road and force themselves across the traffic. This is something which required an enormous number of people on bicycles to achieve. Bicycles were enormous in China, but there has been an enormous decline and that country is now famous for its massive traffic jams.

Cycling declined in both the UK and the Netherlands until
the late 1970s. The Netherlands (top line) reversed the
trend while the UK did not.

Finally, the Netherlands also provides a good example of decline from a high base.

Just like the UK, the Netherlands also saw declining cycling from the 1950s until the mid 1970s as roads were redesigned to accommodate more and more motor vehicles at the expense of cyclists. The UK and the Netherlands followed a very similar decline from the 1950s through to the 1970s (though the decline in the UK reached a deeper point).

Though the Netherlands now has the highest cycling modal share in the world, this country has actually still not grown all the way back out of that enormous decline.

Growing cycling is a slow process, even here. It required an enormous amount of work over many years to achieve a relatively small rise from the dip of the 1970s. The trend to higher cycling levels only came after a second revolution took place on the streets of the Netherlands. Despite a high level of investment, further progress has been slow for the last 20 years.

Just as growth is a slow progress, decline also takes time. The slow rate at which declines occur can hide serious issues of under-investment and bad planning for many years, especially if other factors (e.g. growth due to demographics (students / older people) hides the decline. The Netherlands needs to learn from its own past and take the warning from Denmark (see above) seriously. Cycling is no more "in the blood" of the Dutch than it is of the Danes and cycling will just as easily go into decline here if cycling conditions are made steadily worse. In some places this has already happened due to such things as shared space scaring people off their bikes and other similar mistakes.

If subjective safety is no longer taken seriously enough in the Netherlands then people will stop cycling, the decline which began in the 1950s will continue, and the period between 1970 and 2010 will appear as a difficult to explain plateau on a future version of the graph above.

Assen in the 1970s. Many streets in other countries still look this. If Assen still looked like this then there would not be so many people cycling as there are now. Watch a video of how this street looks now. The traffic lights were removed long ago, but the through traffic went first. Unfortunately, such lessons have been forgotten, leading to new examples poorly designed infrastructure.

So what happened to the "tipping point" ?
From the above examples we can see quite easily that merely having a high cycling modal share is not enough to ensure that cycling continues to grow. Many examples exist of places which have or had a much higher cycling modal share than Birmingham's target of 5% of journeys by bike, but which which have since gone into decline and continued to decline for long periods of time.

All that is required to cause cycling to drop is for conditions conducive to cycling to disappear.

When cycling becomes less safe, less pleasant, less convenient than it used to be, people will switch to other modes of transport.

A bicycle tunnel under the railway
in Assen was constructed in the
1970s with a 5% incline. This is
now considered to be too steep.
The Dutch population is ageing
and requires an ever improving
quality of cycling infrastructure
merely for cycling to stand still
Without improvements, the city's
cycling modal share could drop.

How can we stop this from happening ? Constant new investment and ever improving conditions for cycling. Cycling must be safe, attractive, pleasant and efficient as a means of transport to all locations. Conditions must be such that everyone wants to cycle, not just a small section of the population, because if cycling is only for a brave few then the modal share can only mirror the small segment of the population who cycles.

Because there is no low level of cycling which will grow automatically, asking for little and expecting to achieve much makes no sense at all. Real world results are proportional to countries' expenditures.

The Netherlands spends €30 per person per year on cycling infrastructure even after 40 years of effort in building the required comprehensive network of routes because there is no choice but to do this, because the alternative is to watch cycling decline. However it's important to note that this higher level of expenditure than any other nation doesn't really cost anything. While badly designed and constructed cycling infrastructure costs money and gives few benefits, the benefits due to good cycling infrastructure are greater than the cost.

Remember that even the Netherlands has not yet grown back out of the decline between the 1950s and the 1970s. Denmark's troubles with cycling should be seen as a particularly strong warning to this country. If the Netherlands copies from Denmark then it could very easily suffer the same decline as has Denmark.

But while the Netherlands has no choice but to go it alone and continue to try to maintain the lead, other nations do have a very clear example to follow. The Netherlands is the most successful nation in cycling and it is therefore where the best solutions are likely to come from. However, mistakes have been made even here and it sis important to take inspiration from the very best examples. On this blog we try to help by providing examples of what works and examples of what not to do. We also run regular study tours on which these concepts are demonstrated.

A note about demographics
Locations with universities generally have more cycling than other similar locations. Areas which become the new trendy place for young single people to live (i.e. where a process of "gentrification" or an influx of "hipsters" has been seen) will often see an upturn in cycling. Neither of these things is due to the infrastructure, they are due to the average member of the population being easier to attract to cycling because these demographics are less likely to be put off by those things which would put off other people from cycling.

Demographic factors are always important. Not only infrastructure but also the people that are served by it as well as other factors such as the geography make a difference to the potential of any given location.

The best infrastructure allows any location to fulfill its full potential, whatever that potential might be. Groningen currently has three times as much cycling as Cambridge. However if Cambridge had the infrastructure of Groningen then it might well achieve a higher cycling modal share due to the helpful combination of more favourable demographics, local by-laws banning student cars and milder weather.

At present, Groningen is making far better use of its potential for cycling given other factors, while Cambridge is not.

Update July 2014
I've been writing about the decline in cycling in Denmark for six years and after years of denial, it seems that at last some people in Denmark have started to talk about it as well. This is very good news for Denmark. It is only by recognizing a problem that it can be fixed. Publicity alone does not grow cycling. Pretty pictures don't do it, and nor does international marketing. It takes infrastructural change to encourage people onto bikes. Journeys by bike need to be made safe and convenient.

Update 2015
Read a new blog post about how when it was already in decline in New Zealand, planners ignored cycling and allowed it to wither.

Click here for more posts about statistics and cycling

This post was started some months ago but I finished it today after reading an excellent post on aseasyasridingabike which makes a very similar point about the idea of a "critical mass". Also interesting this week is ibikelondon's piece about the decline of cycling in China. Also read a Crap Cycling and Walking in Waltham Forest blog post from 2011 on a similar subject.

A temporary signalled crossing for cyclists (Road works vs. the Dutch cyclist)

Where cycling is at a high level, chances can't be taken with what would happen if cyclists were redirected from safe and direct cycle-paths onto busy unpleasant roads. The result could be a catastrophic change in how people travel.

Cycling is very fragile.

If people have unpleasant experiences when cycling, they are likely to give up and make their future journeys by a different mode. This is why a comprehensive grid of very high quality infrastructure which keeps cyclists away from cars is essential to achieve the cycling potential of any area given local demographics and geography. It's also why in the absence of safe cycling infrastructure, no amount of training results in people cycling more.

The details are important. This temporary traffic island makes it possible for people to stop in the middle of the road in safety. In the video, a teenage boy starts to cross the road against a red light and benefits from this island. Note that just as at permanent crossings, cyclists are not required to stop in the middle and wait for another light. Crossing the road is always achieved in one motion.

Though this is a temporary cycle-path, the give way line is marked in an obvious way and there is a very flat transition between the road and the cycle-path. In fact, this temporary cycle-path is smoother than the service road to which it connects.

The signs, markings and signals are all familiar to anyone who is used to Dutch cycling infrastructure. There is nothing new to learn, no surprises due to anything operating in an unusual way. This consistency leads to safety.

In this instance it was not possible to construct a temporary crossing right next to the old crossing. Therefore the redirection signs had to be clear and obvious. This sign design is standard for redirections during road-works.

This is not an example of something exceptional, but of the norm. In the Netherlands it is normal that cyclists' routes are maintained when cycle-paths have to be dug up. Please read the many other examples of road works vs. the Dutch cyclist going back six years on this blog or see other examples of maintenance.

Challenges to growth in cycling in the Netherlands. Older people cycle more than ever before but their injury rate is rising.

One graph can never show the multitude of factors behind the cycling modal share of a nation.

In this post, I'll point out how the usual chart which shows how cycling has grown in the Netherlands since the 1970s hides some major demographic factors which influence cycling, not all of them positive.

In addition, we've recently been told that for the first time in nearly ten years, cyclist deaths rose in the Netherlands last year compared with the year before. What's this all about, and is it somehow related to the modal share ? Read on...

First challenge: Immigration
Each year, the Netherlands loses about 30000 more people due to emigration than arrive as new immigrants. Many people who emigrate were born overseas and are returning to their land of birth, but also many Dutch people emigrate. This number, you'll notice, is smaller than that of births vs. deaths.

Nevertheless, by 2010, this exchange of people between countries resulted in 11% of people living in the Netherlands having been born in other countries. Because all other countries have a lower cycling modal share than the Netherlands, this of course implies that immigrants were born in countries where it is normal to cycle less than in the Netherlands.

Immigrants to the Netherlands try hard to integrate well with the local population and this includes their cycling habits. Immigrants living in the Netherlands cycle about as much as they would had they been born in and still lived in Finland or Germany. i.e. less on average than the native Dutch, but as much as people from other leading cycling nations. Clearly, though, we can't state that growth in cycling is due to immigration.

Second challenge: Ageing
Between the 1970s and now, the population of the Netherlands grew from 13 million to 16.5 million people. As we've seen, this is not due to immigration. It is also not due to a high birth rate. As is normal for developed nations, the Total Fertility Rate of the Netherlands (how many children are born on average to each woman) has been less than 2 for some time. In the Netherlands, the TFR has been less than two since 1973.

Births still outnumber deaths in the Netherlands by almost 50000 people per year, and this is what has lead to population growth. Deaths are lagging behind births because of the increase in life expectancy and resulting increase in the average age of the population. Men can now expect to live 7 years longer than in the early 1970s and women 5 years longer. Women still live slightly longer on average than do men.

Meeting these challenges
Neither older people nor recent immigrants from countries with lower cycling rates are the easiest people to convince to cycle.

Immigrants are encouraged to cycle. A successful initiative called Fiets Vriendinnen ("Cycling Girlfriends") is the only organisation that I'm aware of which cycle training for adults. Specifically, they offer cycle training and support for women new to the Netherlands, performing a useful function of helping people to integrate into Dutch society.

However, something else has happened which many people may find surprising. Elderly people in the Netherlands now cycle nearly three times as much as they did in the 1980s. It is the rise in the number of retired people cycling which has driven the growth in electric bicycle sales in the Netherlands.

A group of older people who passed us a couple of Sunday afternoons ago. It is not at all unusual to see groups of retired people cycling for pleasure in the countryside.

Other growth has come from specific areas. For instance, cycling to railway stations, where literally hundreds of thousands of extra cycle parking spaces have been built in an attempt to keep up with growth in usage in the last few years (40% of train passengers now arrive at railway stations by bicycle). Many Dutch towns have seen increases in cycling alongside improvements to infrastructure. This is true of the capital, Amsterdam, of the top cycling city in the Netherlands, Groningen, and also of the small town of Assen, where we live.

What about that rise in transport related deaths ?
Obviously any rise in deaths is something that has to be taken seriously. First some context. Overall, roads and cycle-paths in the Netherlands are amongst the safest in the world. They are much safer than they were in the early 1970s.

The improvement in safety have come in large part from following the principles of sustainable safety. Different modes are separated, keeping cyclists away from the danger from motorized vehicles even when you might not think this was happening.

While 2011 was a slight increase over 2010, this is clearly a
downward trend over time. It is not an imaginary trend.

Sadly, for the first time in several years, 2011 saw the number of deaths on Dutch roads increase. This has seen alarming commentary from some quarters about how the roads are becoming more dangerous, but I do not believe this to be the case. Rather, this is a bump on a graph which still depicts a downward trend. The rise was small, 3.3%, and 2011 was still safer than 2009. Perhaps the same sorts of comments concerning a "rise" were made when this last happened, in 2003.

In 2011, there were 661 deaths compared with 640 in 2010 (these figures are for all road users, not just cyclists). December 2011 was a particularly deadly month with 81 people losing their lives in that month alone. Without this unusual one month peak, the figures could have been lower than for 2010.

There is always "noise" in figures of this nature. It's important not to take too much notice of year on year changes as they can be so misleading. While one cannot make any prediction based on the graph, it looks more like the continuation of a steady downward trend than the beginning of an increasing trend, and we should note that the December figures which particularly made a difference last year are likely to be due to particularly unhelpful weather.

Demographics of road casualties
Of the 661 deaths last year, 40% or 269 were of people over the age of 65. Nearly 1 in five, 126 people, were aged over 80. This is a reflection of how active older people remain the Netherlands and also of how vulnerable older people can be in crashes which might cause less serious injuries in younger people.

We see the same if we look just at cycling casualties.

Over 65s cycle around 12% of the total kilometres cycled each year in the Netherlands. However, almost 2/3rds of the cyclists who died in the Netherlands last year were of people aged over 65.

In total, 200 cyclists died in the Netherlands last year, a rise of 38 over the year before. This is quite a sharp rise in one year. It compares with 185 in 2009, which makes for a less dramatic change. It is still part of a downward trend, and the Netherlands remains the safest country in the world for cyclists.

Why are older people falling victim more often ?
It is an unfortunate fact that as you get older, you also become more delicate. Injuries which might cause nothing more than a little discomfort when you are young can cause a broken bone when you are older. Those which might have resulted in a broken bone can result in death.

The rise in electric bicycles has been blamed for older people suffering more injuries and deaths. One study claims that while electric bikes are not much different in safety to non assisted bikes for younger people, there is a relationship between an assisted bike and a higher rate of injury amongst older people. Such bicycles are sold almost exclusively to retired people, and for this age-group they result in being able to ride faster and further than before, without improving the rider's reaction time or strength.

New figures from 2013 show that women over 60 with
electric assisted bikes (solid red line) are especially
at risk of injury. Two possible explanations are offered.
1. More fragile people tend to buy electric bikes.
2. Electric bikes lead to higher danger for this group
In either case, extra investigation is needed.

However, it would appear that the main reason why older people are falling victim more often is that they are cycling more often. Between 2000 and 2010, the number of kilometres ridden each year by over 65s more than doubled. Over 65s now cycle nearly three times so much as they did in the 1980s.

In the Netherlands, cycling suffers from the "Golf effect". Just as regular gentle exercise by cycling extends life-expectancy, golf has been shown to do the same. However, a surprisingly high number of people die of heart attacks when playing golf. It's not the game that kills them, as golf has few risks, however the combination of an aged demographic and that playing golf can take quite a long time means that there is a fair risk that regular players will die when they are playing golf. Many older Dutch people spend a considerable amount of time cycling, and the same will be observed.

None of this is new
In 2005, a study showed that the number of visits to hospital by older cyclists had risen by 55% between 2005 and 2009. The total distance cycled by this group had also risen by 50% over the same period.

In 2010, a study showed that elderly cyclists ran 3.2 times the risk of injury than younger cyclists, and that this was due not to having more crashes, but due to the relative fragility of older people. It also showed that 10% of total cycled kilometres in the Netherlands are ridden by the over 65s, and a link was shown (again) between improvements to cycling infrastructure, an increased rate of cycling, and lower risks for all cyclists.

What about the future ?

Average number of bike rides per day for Dutch
population separated into different age groups

It is expected that the ageing population will eventually reduce the cycling modal share because older people in the Netherlands cycle less than younger people.

In 2007, a study suggested that ageing of the population could be expected to slowly reduce the rate of cycling in the Netherlands. It also predicted a rise in serious bicycle crashes of 10% over 20 years for much the reasons that we have seen start to occur. This study also showed that immigration was thought unlikely to have a very large effect because the number of immigrants remains small.

That same study includes an interesting table which shows the percentage of cycle trips by over 65s in 2005 for each local government area, compared to what it is expected to be by 2025. In Assen, for example, 12% of cycle journeys were made by over 65s in 2005, and this is expected to grow to 16% of cycle journeys being made by over 65s in 2025. This is about average for the whole country.

Of course, the Netherlands isn't standing still. A new organisation, the National Investigation into Cycling Safety has begun to work on improving future safety for cyclists in the Netherlands.

It is important that standards for cycling infrastructure continue to improve. Only with improvements in standards can the cycling modal share be preserved despite the challenges of demographic change.

All three types of safety are important.

About comparing sensible figures
Last week, British politicians compared Dutch and British cycle safety figures in rather a confusing and perhaps even deceitful manner by referring only to the absolute number of deaths in both countries.

It is of course true that the Netherlands has a higher rate of cycling injury per hundred thousand people. However, this is because while everyone cycles in the Netherlands, and as we see above, this covers a very wide range of demographic groups, only a minority cycle in the UK and they're largely young adult males.

Comparing per km cycled, even though the Dutch cyclists include far higher levels of more vulnerable groups (not only older people, but also children and people with disabilities), the figures are very much better for the Netherlands than the UK, even taking into account the increase last year. Several people blogged about this, but I refer you to "(Drawing) rings around the world" for accurate figures.

Someone quipped that by the same logic as used by the British Transport Ministers, Switzerland could perhaps learn a lot from the UK's good record on skiing safety, and it's a valid comparison.

Any British politician who wants to know what the difference really is between cycling in the Netherlands and cycling in the UK only has to ask us to find out. We do tours specifically for this reason, but not one British politician has yet come to see for themselves.

2014 update
Traffic safety figures in the Netherlands have continued to improve despite the upward tick in 2011. The total number of deaths in 2012 was 650, roughly the same as 2011 but this improved in 2013 to 570. That's the lowest figure ever. 184 cyclists lost their lives in 2013, only five of whom were under 15 years old. 124 of the cyclist deaths were of people aged over 60. Over 2/3rds of the total and  continuing the trend of their being the most vulnerable people being injured in single-sided crashes.

Investigations into the cause of the rise in injuries and deaths continue. If the person who made the quip about skiing safety reads this, let me know and I'll link to you.

"The busiest cycle street in the world" ? Why do we keep hearing exaggerations about Norrebrogade / Queen Louise's Bridge in Copenhagen?

An industry seems to have grown up around promoting Copenhagen as the world's top cycling city. Sadly, some of the claims made do not stand up to much scrutiny.

For instance, Nørrebrogade is claimed to be "the busiest cycle street in the world". It's certainly the busiest cycle street in Copenhagen. Being 3 km long and taking a straight line path right through the city, it and the bridge on it are a funnel for cycle-traffic which uses this as one of the most direct routes to several locations. Very many of the photos of cyclists from Copenhagen are shot on this road because it's the busiest.

Counter on Dronning Louises Bro / Queen Louise's Bridge
Norrebrogade Copenhagen. The claims made for counts
of bicycles here do not stand up to scrutiny.

A recent widely reproduced press-release from the city claimed that 36000 cyclists use the road every day. Some websites go further and claim 38000 per day. Usually you have to assume that such printed figures are correct, and that's what I did too. However, in this case someone pointed to me that it's unusually easy to check it for yourself. This count was made on the same street as a famous public bicycle counter which many of you will already have seen photos of. The counter displays both the current date and the number of cyclists who have passed since the 1st of January, so a a simple calculation can be done by dividing one number by the other. Take the photo on the left as an example (click on it to see it in greater size at its owner's website). It shows 1559492 cyclists had passed by the 4th of June 2010, which was the 155th day of the year. That's just 10060 cyclists per day on average by half way through 2010. It is a respectable figure, but not close to the claims so often put forward for the street.

So what's going on here ? I don't know. It's surely good marketing to have "the busiest", but don't we need a bit more than that ? Before Christmas I emailed the two people who were responsible for a document about the street to find out where the numbers came from, but have not had an answer (see below).

Counters in many different places in the Netherlands record higher numbers than this. For instance, eight locations in Utrecht have more than 10000 cyclists per day, and one of them has more than 22000 cyclists per day. In Groningen, six different places have counts greater than 10000 per day, two of those are above 14000 per day. Bear in mind that Copenhagen's population is more than three times that of Utrecht and six times that of Groningen. Counters in larger Dutch cities almost certainly record higher figures.

It's just marketing.


Danish professionals, including the Cycling Embassy of Denmark give a slightly different perspective.

This is the inside back cover of one of their recent publications, giving some numbers. The Danes now cycle for 16% of their journeys. Sadly, this is lower than the 18 or 19% measured just a few years ago and which I have quoted in the past. The difference is significant. It's equivalent to one in eight cyclists giving up. This is not something to celebrate - it's terrible news. We need to know the cause of the decline and work out what can be done about it.

Some of the other figures are also interesting. 36% of Danish adults ride a bike to work "at least once per week". The percentage of daily commutes is not given, but it will of course be lower. 45% of Danish children "often" bike to school. Clearly this isn't the percentage who cycle every day, which again is not given and will be lower. "Often" is open to interpretation but could be less frequent than once per week. This is a startlingly lower rate of school cycling than is seen in the Netherlands, and that's important. Cycling has to start with children.

Note also the claim that 2/3rds of children under 11 use helmets when cycling in Denmark. This is not something which would be highlighted if it wasn't thought to be important. However, it is again very different from the Netherlands, where despite occasional ill-judged campaigns, helmets are still only rarely seen except on sport cyclists. It's another sign of cycling under pressure. People are scared for the safety of their children, and perhaps this has some connection with the reduced rate of cycling.

Lastly, note that 2.2 metres is the new standard width of cycle lanes in Denmark. That's somewhat narrower than the 2.5 metres which is standard here for single directional cycle paths, and it's again a thing which puts cyclists under a little bit more pressure.

It all adds up to a bit of a murky picture, similar to what I pointed out before. Copenhagen, and in the wider sense, Denmark, has a relative lack of subjective safety and this is stifling bicycle use, even in the top city of the world's second cycling country. Helmet promotion campaigns like this are only likely to contribute to the decline in Denmark.

I'm not happy about any of this. Cycling is extremely fragile, even in the two great cycling nations of the Netherlands and Denmark. It doesn't take much to stop people from cycling. If care is not taken, cycling in the Netherlands could also decline. We need ever better standards for cycle provision and no negative marketing. We all need to be able to learn from what has caused the decline in Denmark.

Why has cycling in Denmark declined since 1992, while it has continued to rise in the Netherlands ? Why is this question not being asked more widely ? This graph ends at the "18%" point. Since that time, the proportion of journeys by bike has dropped further to 16%

It's clear that Danish cycling policies are not succeeding as we all wish they would. This problem of declining cycling has not occurred in the Netherlands, and that is as good a reason as any why The Netherlands remains the best place in the world to turn to if you want to see what successful policies look like.

Update 25/1/2011
There's been quite a lot of discussion in the comments below this blog post of the discrepancy between counts which made up the first part of the post. Copenhagener got close to the answer by finding the cycle counts for a day in September 2009. However, I was also sent very helpful emails from Klaus Grimar, Project Manager, and Soren Zebitz Nielsen, Student Assistant, both of the Center for Trafik (Department of Traffic) in Copenhagen, explaining the numbers. They gave permission to reproduce the first long email, and also a shorter summary which you will find in italics below:

The survey that resulted in the number of 37.000 cyclists on Norrebrogade on the bridge where the counter is placed was conducted manually in early September 2009. We know that early September is a peak period for cyclists with good weather and due to the start of the academic year for the universities in the city. The number of 37.000 cyclists represents the average number of daily cyclist on a normal work day in both directions on the street during the peak season.

As an average in both directions on a normal working day throughout the year, the daily total number of cyclists on the street is around 28.000.

The counter relies on a sensor placed below the asphalt on the bicycle path only. We know that during the rush hour some faster cyclists overtake slower cyclists using the road, and some even the walkway. We find that the counter does not register all cyclists when the bicycle traffic is dense. It is our experience that the automatic counter generally counts around 10% less cyclists than the actual number during the peak hours and around 5% less in normal hours.

The idea of the counter is not to be an official measurement, but more to give the cyclists an idea of the approximate magnitude of bicycle traffic on the street as they pass by and a feeling that each individual cyclist contributes to make Copenhagen a City of Cyclists. The numbers counted by the automatic counter cannot be relied upon for academic purposes.

However there is no doubt that there has been an increase of around 7000 cyclist daily on the street of Norrebrogade as a consequence of the project there, since the manual surveys of bicycle traffic before and after the project were made using the same method and the traffic was counted at same time of the year in the surveys before and after the project was initiated.

Thanks to Klaus and Soren we now know that the widely reported "average" picked up by the marketing people and the media was actually a peak, and we also know the official average for a working day over the year, including a compensation for the cyclists missed by the automated count (other places could perhaps use the same compensation), but still not the average for all 365 days of the year. I'm of course very pleased to see that the improvement in conditions for cyclists on this street resulted in an increase in cycle traffic in this location. However, it has to be viewed alongside the unfortunate decline in cycling overall in Denmark.

See previous articles about Copenhagen. There are also previous articles about helmets, and particularly amongst them, this one which points out that the "danger" of cycling really has been blown out of all proportion next to that of driving. Others have pointed out that a lack of sufficient investment could be behind the lower cycling share in Copenhagen vs. Dutch cities.

September 2012 update


This year, some people I know from Cambridge visited Copenhagen and wrote about their experiences.

They also took some photos, including the one on the right, which while it's supposed to be representative of a street along which "up to 35,000 cycle trips per day" are made, actually shows that something unpleasant has happened on the "busiest cycle street in the world". The photo shows the state of the sign  on the 27th of May 2012. By that date, the 148th day of the year, 1261731 cyclists had passed the sign. In other words, in the first five months of 2012, the number of cyclists riding past this point averaged 8525 per day.

This number is not only a long way short of 35000, but compared with the numbers which I used to make the calculation at the top of this blog post, it's evidence of a 15% decline in cycling in Copenhagen compared with a count from two years previously. Both counts were made by the same permanently installed counter in the same place and both over the first five months of the year. This 15% difference cannot be explained away as an artifact caused by just a few days of bad weather - it's equivalent to 22 days of cycling being lost in just half of one year.

A 15% decline over two years in the busiest street in the city sounds catastrophic. Why has this happened ? Why did Copenhagen choose to hype the cycling rate on this street if it were actually in decline, and why do they continue to hype the amount of cycling in this street in the light of such a decline ? A 15% drop is in line with an acknowledgement by the Cycling Embassy of Denmark last year that cycling had dropped from between 18 and 19% to 16% of journeys within Denmark, but they were referring to a different decline which had happened over a much longer period than two years.

The article from Cambridge also reports a statistic that "Copenhageners make half of their urban trips by bike". However, this has no basis in fact either. We need to start with a little history:

All trips within Copenhagen 2002.
Bicycles "slightly less than one
fifth", cars just over half.

In 2002, "slightly less than one fifth" of journeys within the city were by bike. A target was set in the same year to increase commutes only to 40% by the year 2012. This target was missed. There was some reported growth in that official figures show that commuting by bike rose from 32% in 2002 to 35% in 2010, but instead of acknowledging that the 40% target for 2012 wasn't going to be achieved, Copenhagen instead announced a 50% target for 2015. It would take an extraordinary and unprecedented growth in just a few years to achieve this.

The 50% target sounds great of course, and that is what is supposed to do. It's an artifact of marketing for Copenhagen, not of the real cycling modal share. At best it's only a target, and it's a target not for all journeys but only for commuters. No-one should reproduce this figure as an achieved modal share, because it has never been any such thing. In 2009 when I estimated that Copenhagen's true bicycle modal share was around 22%, the head of Copenhagen's bicycle programme wrote a supportive comment on this blog. If the same rate of growth has been continued for the last three years then we would expect that the true modal share for Copenhagen in 2012 is around 23-24%. However, this is rather difficult to determine because the numbers are no longer easy to find.

All trips within Copenhagen 2010.
33% for bicycles, 40% by car, 27% public
transport. They missed out walking.

While the figures for 2002 (in this document) were truly for all modes, the figures issued for 2010 which appear to show dramatic growth do so in large part because they have omitted to mention the approximately one fifth of all journeys within the city which are made by walking. Assuming that around a fifth of journeys still being by foot, the diagram on the right would indicate that approximately 26% of journeys are now by bike in Copenhagen, which is more in line with my estimate. However, note that these figures are for trips "starting or terminating in" the city and we don't know how this phrase is defined nor how the figures were calculated. We have to be wary of tricks, especially when we're looking at figures from a place with a history of exaggeration.

Any claims for growth in cycling in
Copenhagen have to be judged against the
steady decline in Danish cycling since 1992

People from other countries looking for inspiration are often drawn to Copenhagen because of the deluge of publicity from the city. However, they need to look further than headline figures ("half of their urban trips" etc.) which form part of a marketing effort from Denmark. There was a time when the annual bicycle reports and policy documents from Copenhagen were the envy of the world, and a time when they presented good data which could be compared one year with the next. However, these official documents are no longer what they once were.

The Danish people deserve much better than this. Selling the "cycling city" of Copenhagen as a brand does not help people to cycle in the city. From the evidence that is available it appears that cycling is dropping and not rising, both in Denmark as a whole and on some main streets of Copenhagen if not across the whole city. Less hype, more substance and maybe even a little humility is required before Copenhagen can achieve its potential. More investment is needed to give the city any chance at all of meeting the old target, let alone the new one. The environment needs to be transformed to be truly safe for cycling.

Bad infrastructure designs need to be abandoned and no longer presented to the world as something to copy. In particular, the "Copenhagen left turn" is not, and never was, a good design. Drivers turning right should not be encouraged to merge with cyclists as they approach traffic lights. Advanced stop lines (aka "bike boxes") are not advanced cycling infrastructure - they create conflict which should be avoided and they should not be promoted. Unfortunately these ideas are still being presented to the world as something to emulate.

Not just Copenhagen

These days, it is not only Copenhagen which is doing this. Earlier this year, Amsterdam also presented numbers which excluded pedestrians to make the figures sound better (I wrote about that too). In the last few days I read an impressive claim from someone in Utrecht about a claimed commute only cycling modal share for people who both live and work within the city of around 60% (though the actual cycling modal share for all journeys within the city is apparently around 26%).

Many cities across the world publicize commute only figures. This includes Cambridge in the UK and Portland in the USA. Neither of these cities gets remotely close to the modal share for cycling which is normal in the Netherlands, nor to the wide demographic of cyclists which is not only a curiosity of Dutch cycling but necessary for any place to achieve a high modal share.

Why exaggeration is harmful
Cambridge's campaigners were left wondering how it was that a city where infrastructure "does not seem to be of the same elevated standard found in Assen" should have achieved the high modal share which Copenhagen claims. The answer unfortunately is very simple: they haven't. Rather, they are telling impressive sounding stories which are not necessarily based in fact.

Wherever it is done, playing with statistics in this way harms us all. Without reliable figures, how can any comparison be made between countries or even within the same city on different years ? If facts are obscured by smoke and mirrors, made up figures are shouted from the rooftops, half the audience is distracted by photos of pretty girls on bikes, and a good many people are simply more interested in a nice sounding story than whether what they are being told is factual, then how can anyone tell what works from what does not work.

Extraordinary claims need to be looked into and challenged, wherever they come from. Exaggeration needs to stop.

Update December 2013

A Danish newspaper reports that cycling in Copenhagen has "stagnated" and that the 50% target for 2015 can't possibly be met.

The only way to a higher cycling modal share is to build better infrastructure. We are still willing to help Copenhagen to "Groningenize" itself.

Update 2014
Earlier this year I read reports of how Copenhagen and Denmark in general had learnt from their decline and were changing course. I wrote about this change of direction because of it being extremely good news for Danish cyclists. It is only by recognizing a problem that it can be fixed and it appeared that the Danes had learnt their lesson and were going to fix their problems.

Independent EU statistics for cycling
in Copenhagen for 2010 and 2012.
Cycling dropped from 31% to 26%
Car usage rose from 29% to 33%.

Unfortunately, only a few short weeks after that happened, Copenhagen started fresh claims of having "the busiest cycle street in the world". This time they claim to have over 40000 cyclists per day travelling over the bridge. I'm not going to link to a source because frankly this sort of hype does not deserve any encouragement at all. Copenhageners: Please stop exaggerating your achievements. This is particularly unhelpful when independent figures show a decline between 2010 and 2012.

In any case, a high number of cycles in one spot is not actually a sign of success. It's not something to celebrate. The funneling effect indicates that there are a lack of routes and may actually put people off from cycling.

When similar things happen in the Netherlands, there are other routes which cyclists can be encouraged to use. This is possible because of the high quality grid of cycling infrastructure which covers the whole country. The importance of this has been understood since the 1980s and this is perhaps the most important lesson from the Netherlands yet to be learnt elsewhere.

International marketing of Copenhagen needs to take second place to actually solving the problems facing cyclists in Denmark.

High cycle counts can be indicators of a problem:
A lack of bridges to cross a river forces people to
make detours and creates queues of cyclists. Lots of
bikes in one place may look impressive, but it's a
sign of failure not of success. Don't celebrate
traffic jams for bikes! Read more about funneling.

Update July 2016
The hype over this location and of Copenhagen in general, continues. Five and a half years have passed since I debunked the claim that 37000 cyclists per day ride over the Dronning Louises Bro in Copenhagen and in all that time, no-one has been able to provide a justification for that claim. Rather than attempting to justify the previous claim, the same players have moved on to claim a rise to 42000 cyclists per day passing over the same bridge. Exaggeration doesn't help Copenhagen's cyclists and it doesn't help anyone else either. What is happening is that the facts are being obscured.

What's more, even if the claims were true, they're not necessarily something to be proud of. Thousands of cyclists forced into a detour to take the same route because there are relatively few options available to cross a river may make for nice photographs and good headlines, but it does not make for efficient cycling.

Funneling is bad for cycling. It's better that cyclists can choose from a finely spaced grid of high quality routes which enable each person to choose a relatively lightly trafficked and more direct route to their destination. That, though, is of less use if your primary goal is to market your city internationally by use of impressive looking photos and ever more exaggerated figures.

No-one celebrates traffic jams for cars. They shouldn't celebrate traffic jams for bikes either. Big queues of bicycles are a sign of failure. A comprehensive cycling grid allows more efficient journeys to be made.

Pit Canaries

Seen many cyclists recently ? There is a huge difference between the rates of cycling in different countries.

The graph on the left shows the rate of cycling as a percentage of all journeys made in sixteen different countries. This is also known as a modal share or modal split. Australia, the USA and the UK are at the left with 1% of all journeys by bike, while Finland, Denmark and the Netherlands are on the right with 11%, 18% and 27% of journeys by bike respectively.

There is a huge difference between the numbers of cyclists you see in different countries. There is also a huge difference between their behaviours.

So, back to the question. How many cyclists do you see when you go out, and how many of them are dressed in a way that indicates that they feel they have to worry about their safety on the roads ?

Are your cyclists a small part of traffic, wearing helmets, dressed in fluorescent jackets and predominantly young and/or "sporty" ? Do your cyclists cycle "vehicularly" and identify themselves as "cyclists"? Or do you live in a location where cyclists are of all ages, both sexes, and generally ride in normal clothes with no worries about visibility ?

Cyclists are the pit-canaries of the roads. If they're numerous, dressed in ordinary clothing and wide-ranging in age you can tell that you are in a location where cycling is "normal" in society and where it is safe enough, and feels subjectively safe enough, that everyone cycles. If people feel they have to dress to be safe then this is a sign that they do not have adequate subjective safety.

It goes further. Do you see women routinely cycling alone at night ? Do children cycle long distances alone, day and night ? Do most children go to school by bike ? Do parents consider it to be safe for their children to do this ? Do those parents who do let their children cycle make them wear helmets and fluorescent clothing ? These are indicators of how subjectively and socially safe people feel and are a determinant of health of cycling in an area.

Make cycling a truly attractive thing to do and more people will do it. Here in Assen, children are regularly seen riding great distances with groups of friends. Some cycle 20 km to school in the morning and 20 km back again each evening. Take a look at a video of how children get to primary school here.

As mentioned a few days ago in "Reclaiming the streets", cyclists are also the pit-canaries of society. Societal problems can also lead to low rates of cycling. If there is a significant risk of mugging (or worse) in your area, you will see fewer cyclists. This is a problem for social safety. It is surely no co-incidence that the three countries with the highest cycling rates also have progressive social policies. You will find a very similar ranking of countries in the results of a UNICEF report summarized by the BBC with the title of "Why are Dutch children so happy ?".


Dutch infrastructure works so well to convince people to cycle that recent immigrants to this country cycle on average about as much as people who live in Finland.


The graph at the top is from John Pucher and Ralph Buehler's article "How Cycling was made Irresistible" - one of many articles referenced on our cycling articles webpage. The Dutch really have made cycling into an irresistible thing to do. Everyone cycles for at least some of their journeys.

We run cycling study tours here in the Netherlands on which we demonstrate the how it is that the Dutch have achieved their amazing high cycling modal share. We also run cycling holidays in this location, which we believe is the most pleasant in the world for cycling.

Sadly, since this blog post was originally written, the cycling rate in Denmark appears to have dropped quite sharply.