[Up: Table of contents]
[Previous: Record of dialog]
[Next: Cambridge city government actions]

Descriptions of Dutch and Danish
cycling and design standards
without rose-colored glasses

The proponents of the Vassar Street project point to Danish and Dutch design standards as their model, so it might be helpful to gain some insight into those standards from people who are concerned about them. It is also useful to point out that the Germans, having read -- and generated -- research documents that reveal the problems with sidepath systems, have taken a decided turn away from them, toward accommodating bicycles on streets. [Details]

John Forester, from his book Bicycle Transportation (1994):

The VeloCity conferences have been organized in Europe for over a decade. They are conferences of bicycle planners and bicycle activists. Because the 1992 conference [in Montréal, Québec] was the first of these to be held in the Western hemisphere, it provided the first chance for North Americans to meet and discuss cycling transportation engineering with many of those holding responsible positions in Europe...The dean of traffic engineering at one of the three schools in Holland that teach traffic engineering, and researchers in charge of the largest current bicycle planning research project in Denmark, did not even understand (although they are fluent in English -- language was not the problem) the cycling-traffic-engineering questions that Americans asked of them. The questions that had been debated and investigated in the U.S.A. for two decades...were so far removed from their frames of reference that they didn't understand them. The Dutch dean of traffic engineering was asked to describe the principles and data upon which Dutch traffic engineers based their bikeway designs. Once he grasped the significance of the question (which took several minutes of discussion in itself), he said that they had none, that they just used "common sense." The Danish researchers were even more illuminating. They showed slides of their latest research project. One showed a cyclist using, as the researcher specifically said, the facilities as they were meant to be used. The picture showed a cyclist turning left from the bike lane directly in front of an overtaking bus. That shook us Americans. There was no doubt that the cyclist was moving: she was leaned over for the turn with both feet on the pedals while the headlights of the bus loomed over her shoulders. They showed other slides of their attempts to prevent motorist-right-turn car-bike collisions at intersections that incorporated bicycle sidepaths. They did things like changing the height of the sidewalk, first lowering the curb to let traffic mix but later installing berms to separate it again. They had no idea that, in topological and and traffic-engineering terms, the berms were identical to the curb, and that the problem was that they were putting straight-through cyclists on the right of right-turning motorists. They showed slides of their efforts to prevent accidents to cyclists at bus stops. The problem was that, like so many other places in Europe, they put cyclists on the right- hand side of unloading buses. Naturally they created bike-pedestrian collisions because the descending bus passengers couldn't see the cyclists coming. Even after they had been told (by me), they had no idea of what they were doing wrong. They had no knowledge or recognition of the idea that cyclists are drivers of vehicles and are constrained by the physical and mental laws that control the behavior of vehicles and their drivers. Of course, if your experience has been with cyclists who amble along at 6 miles per hour or less, for short distances, and often while mixed up with pedestrians, it is easy to lose sight of the fact that cyclists operating in a more useful and efficient manner are drivers of vehicles instead of rolling pedestrians. That is what these Europeans (not all Europeans) have done. They have designed for cyclists as rolling pedestrians and have largely got away with that error, even being praised fulsomely by those who don't understand.

Recent Changes in European Opinion


In Germany (a place where law and order prevails and people tend do as they are told), the German Cycling Club (ADFC) now advocates bike lanes. It is not that bike lanes are particularly beneficial - they haven't yet really understood the problems with bike lanes -- but that they now know that any anything, even bike lanes, is better than the mandatory bike paths that government has imposed on them for decades and they had been accepting without question.

Even the German government is changing. In response to the opposition of the cycling organization to its bikepaths, in 1992 the Berlin Government paid for a study to prove that its bikepaths were safe. However, the mathematics of the study were erroneous, as the Berlin traffic engineers demonstrated in a series of letters, and the study actually showed that the bikepaths were more dangerous than normal streets. The German road research organization then ordered from the University of Hannover a study of safer intersections (Safeguarding of Cyclists at Inner-City Intersections [Sicherung von Radfahrern an städlichen Knotenpunkten]). The safer intersections with bike paths were still more dangerous than normal street intersections. The Christian Socialist Party didn't like the conclusions of these studies, so it ordered a study by the German insurance organization. Its study (Accidents to Cyclists in Bavaria) appears to be anti-cyclist in tone but it still concludes that urban bikepaths are more dangerous than normal streets.

Bicycle Transportation may be obtained from MIT Press.

Jeremy Parker, in an e-mail message:

Amsterdam is the only place I have seen worse than the USA for cyclists running red lights, and mopeds do it too. The system couldn't function without that.

Dutch cyclists are allowed to have one dog alongside on a lead, but not two dogs.

The Netherlands divides vehicles into two kinds, slow vehicles and fast vehicles, but I'm not sure of all the implications of this. I think mopeds and farm carts, and maybe tractors, count as slow vehicles along with bikes. Slow vehicles are banned from great chunks of the road system. There's more to the classification than this, though. From the time that the Germans were occupying the Netherlands in WW II, until very recently, a bike had to yield to everything at an uncontrolled intersection, no matter what its direction. Presumably that meant that in an accident the cyclist was always in the wrong. I think the law only applied to bikes, not other kinds of slow vehicles, and anyway I think the law has now, at last, been repealed.

Not every kind of slow vehicle is allowed on bike paths. There's a maximum permitted vehicle width. I don't know what the width is, but I remember that when Denmark was thinking of introducing a similar law, there was a big fuss, because it would have resulted in banning a number of types of recumbent tricycles. The Danish width and the Dutch width might not have been the same, however.

The distinction between slow vehicles and fast vehicles might also be breaking down because of mopeds. In a pilot project a few years ago it was found that permitting mopeds not to use bike paths resulted in cutting two thirds off their accident rate. The result was turfing mopeds off urban bikepaths. I don't think there is any movement to permit a similar experiment for bikes, however.

I can talk about Amsterdam, but would like to point out that here in England we have Cambridge, with a higher bike modal split than Amsterdam. We've got bike friendlier cities too, including three quite large "new towns", Stevenage, Harlow and Milton Keynes, where the bike path networks were built first, and then the towns built round them, resulting in much better networks than you could ever get by trying to retrofit a network into an existing town. It was Stevenage that taught the Dutch how to do bike networks, after they had forgotten about bikes for 21 years, between 1952 and the 1973 oil cutoff by the Arabs.

The main thing about Amsterdam is how small it is. You could walk from the main train station, which marks the city center, to the ring road, their beltway, in about an hour, if you pushed it - it's only about 4 miles. The Dutch reckon their bikes are about three times as fast as walking, so you can get anywhere by bike pretty quickly (British or US bikes are 4-5 times as fast as walking).

Amsterdam carries segregation about as far as it can go. As well as networks for pedestrians, for cars etc., and for bikes, there are two other segregated networks, one for trams and buses, and one for boats. I suppose it was the cars that came last, and they do seem to get the worst deal. I haven't tried driving in Amsterdam, and don't want to, even though there is a network of freeways approaching it from all directions.

With all those networks, and only the boats having bridges to truly separate them, it's a problem to keep segregation where one route crosses another. Hence all the red light running by the cyclists. During my time in Amsterdam I left my bike in the hotel room most of the time - I wasn't going to trust it to the hotel rack - and walked. The Netherlands does have one corner that isn't flat, Maastricht, down by the Ardennes of WW II Battle of the Bulge fame. Cycling rates there are not as high as in the rest of the country.

Forester replies to Parker:

Jeremy Parker's report of VeloCity in Amsterdam this year [2000] replicates my information from VeloCity in Montreal in 1992. At that time it was obvious that we Americans had produced far better information about bicycle transportation engineering, because we were operating in an environment in which it was possible to fight off the unscientific bikeway mania. We showed, first, that there is a sound scientific basis (as scientific as traffic engineering ever gets) for the vehicular-cycling principle, and, second, that there was no scientific basis of any kind for the bikeway hypothesis, which, therefore, was nothing more than a superstition. The additional information developed at each of these VeloCity conferences merely confirmed the absence of any scientific basis for the bikeway systems, because those in charge of designing the systems were unable to produce any such basis.

When those in charge of designing a system cannot point to any scientific basis for that design, then it is absurd for anyone else, be he bicycle advocate, bikeway advocate, urban planner, or whatever, to praise the rationality of the system and its ability to produce some wished-for result. If a system is specifically designed on a scientific basis to achieve some desired result, then it is reasonable to praise that system if it does achieve that desired result. If, on the other hand, a system is created without any scientific basis for its design, as was the Dutch bikeway system, then whatever may be its results, they are purely fortuitous. No, that's not quite correct. It is reasonable to conclude that the Dutch bikeway system was designed to get bicycles out of the path of overtaking motor traffic, to "separate" them, to use the conventionally inaccurate term. As we can see, it does that quite well. However, to suppose that such a system would produce other wished-for effects, such as producing safer and more efficient bicycle transportation, is absurd because there are no grounds for believing that there was ever any scientific information (information which, if it exists, has never been presented) to serve as the basis for designing any such effects into the system. Such results, therefore, must be fortuitous. Therefore, if such effects are observed, they are most probably produced by some other causes than the bikeway system.

To maintain that some wished-for conditions have been produced by the Dutch sidepath bikeway system requires close analysis. First, the wished-for conditions must be demonstrated to exist in conjunction with the bikeway system. If such conditions exist, they may be produced by a variety of causes. To merely assert that the bikeway system is the cause is logically invalid, because there are no grounds for believing that the bikeway system designers had any scientific basis for designing to achieve such a result. One wished-for result is that the bikeway system makes bicycle travel more convenient, enables cyclists to reach their destinations in less time than would normal travel on the road. To maintain this claim requires both analysis of the routes available by bikeway and by road, and analysis of the speeds and delays along each route. Such an analysis shows that, in general, bikeway routes are longer than road routes (because not all roads have bikeways), that the speed of travel along paths is no greater than that on the road, and often is less (on roads, when moving, cyclists can usually travel as fast as their legs will drive them, while on bike paths the traffic limits safe speed to a lower level), that the number and total time of delays is less on the road than on urban sidepaths (where every intersection and many driveways impose slower speeds, more probable delays, and longer delays than does the road). In short, this analysis demonstrates that, in fact, urban sidepath systems on the Dutch model decrease the convenience of bicycle transportation, by requiring probably longer routes at slower speeds with more delays. This shows that because the Dutch bikeway designers did not design to, at least, maintain the convenience of bicycle travel, they made it less convenient. That is one of the fortuitous results of designing without a scientific basis.

Now consider the big claim for Dutch bikeways, that they reduce the cyclist accident rate. It can be shown that this design reduces the proportion of higher-speed collisions between cyclists and motorists. This is probably the basis for the claim that the Dutch cycling fatality rate is low. However, analysis of traffic movements, car-bike collision statistics from Holland and Denmark, and one experiment in the USA (so dangerous that it has never been repeated) show that if the sidepath system were used at normal road cycling speeds, the rate of lower-speed car-bike collisions would be greatly increased (the factor derived from the only experiment that has been dared was 1,000 times). The Dutch system does not show such an increase. The reasons are obvious: the Dutch don't ride like cyclists on the road, they ride like rolling pedestrians, and they suffer many delays both from caution and from the traffic-control devices that have been installed to protect them by preventing them from riding like road cyclists. Does this mean that the Dutch bikeway system has reduced the cyclist accident rate? That's an iffy question. The Dutch produced a very dangerous bikeway system, compared to cycling on the road, but they have overcompensated for those dangers by installing protective measures that make it extremely inconvenient, again compared to cycling on the road.

One could consider this to be a trade-off that can only be judged by taste. That is, which does one prefer, slow cycling at an accident rate that is less than normal, or fast cycling at the normal accident rate? However, there is another course that can be taken. That is, improve traffic-cycling skill, to achieve fast cycling with an accident rate about 25% of that of the typical cyclist. Surely, that is the much better solution. That is the solution taken by well-informed cyclists in the USA, and some other nations also. We prefer to ride competently, lawfully, and conveniently with our present rate of accidents, over being forced to ride much slower by a system that is so dangerous that it becomes reasonably safe, and maybe not even as safe as our present cycling, only at half or less of our present average speeds. Given reasonable competence, easily attainable by nearly all people (and certainly by all people who could be safe in a bikeway nation), the roads accommodate all speeds of cyclist in reasonable safety, while the Dutch system forces all cyclists to ride very slowly for their own safety, and to avoid prosecution for attempting to ride in a convenient and safe manner.

Riley Geary

[Who maintains a Web site analyzing crash data from the Fatal Accident Reporting System and other sources.]

I'm sure this won't come as much of a surprise, but they [Dutch officials and academicians] still didn't offer any concrete data to back up their claims when pressed on this point at the Amsterdam Vélo Mondial conference last month [in 2000]. When I asked the program director, who is an official with the Dutch transport ministry and was a panelist in the "Integration or Segregation" workshop described previously by Jim Baross and Jeremy Parker, about this privately -- after dodging around the issue some more he finally came up with the amazing pronouncement that he really wasn't interested in statistics or looking too closely at the actual safety record of their various facilities!

At least I knew then why my own presentation on US bicycle fatalities based on FARS data had been downgraded to a mere poster session, and why there were virtually no other presentations offered involving any real traffic safety data through either the workshops or other poster sessions (David Tomlinson's presentation on a decade's worth of car-bike collisions in Toronto being the single noteworthy exception).

The Dutch bike planners clearly think they have all the answers (which I would paraphrase as "Segregation today, segregation forever!"--these are after all the closest relatives of the people who originally coined the term "Apartheid", even if they are far more socially tolerant than their Afrikaaner cousins), and while they're more than happy to share their vision of how they think things ought to be with those of us living in less "enlightened" or "bicycle-friendly" societies, just don't try and confuse them with any inconvenient facts or ask too many awkward questions about safety and freedom of choice --they're far too smug and condescending to entertain such potentially heretical thoughts.

Riley Geary
Arlington, VA

[Up: Table of contents]
[Previous: Record of dialog]
[Next: Cambridge city government actions]