Defects of the Design-Cyclist Approach
as Adopted by the AASHTO
Guide for the Development of Bicycle Facilities

home                 government

1 Introduction

This report is a review of the design-cyclist concept that is formally introduced in the draft edition of the AASHTO Guide for the Development of Bicycle Facilities that is marked Draft, April 1995. This report was first intended to concern only the newly introduced description of the cyclists for whom the bikeway system is designed, the design cyclists. However, this new formalization of the design concepts that have always underlain the former editions of the Guide enables rational criticism of the Guide's actual words, rather than having to base criticism on imputed intents. Therefore, while the following analysis is based entirely on the Guide's section about design cyclists, the analysis covers many of the concepts on which the Guide is based.

I intend to provide other comments later.

Some history and a few basic principles are necessary for understanding the basis for the bikeway controversy.

2 The Bikeway Standards Do Not Reduce Accidents to Cyclists and Do Not Reduce the Level of Skill Required

The AASHTO Guide sprang from the second California standards, the ones that were prepared by the California Bicycle Facilities Committee. The first California standards were prepared, by order of the Legislature, by UCLA in 1972 and called Bikeway Planning Criteria and Guidelines. I have a copy; I don't know whether others do. They were prepared without consulting or notifying cyclists. After these were delivered, the California Legislature established the California Statewide Bicycle Committee to review the California Vehicle Code, also without informing cyclists. I discovered that this committee had met and I attended the second meeting, at which I talked my way onto that committee. During the operation of that committee I discovered that its purpose was to recommend to the Legislature laws that would prohibit cyclists from using roadways that were near bikeways. Later, I discovered the existence of the standards for those bikeways that government intended to build. So far as I could discover, the committee's existence and operation were sparked by the California Highway Patrol and the Automobile Club of Southern California. I prevented the enactment of the mandatory bike-path law. I could not prevent the enactment of the mandatory bike-lane law. However, I did prevent adoption of Bikeway Planning Criteria and Guidelines by pointing out the extreme dangers to cyclists of the designs therein. Others may have had a part in that rejection, but I was the first to analyze the dangers and the only one to publish my criticisms.[1]

My analysis was the first scientific study of the movements of cyclists in traffic. It showed that cyclists should behave like all other drivers, and that when they behaved in the ways required by bikeways they became endangered.

Also, during the operation of the California Statewide Bicycle Committee, Ken Cross presented the first scientific study of car-bike collisions, those in Santa Barbara County. This study was funded by the California Office of Traffic Safety. This study showed that nearly all car-bike collisions were caused by turning and crossing movements and very few by straight-ahead motorists hitting straight-ahead cyclists, which are the only kind of accidents that bikeways are intended to, or can, prevent. At Ken Cross's presentation to the committee and others, I immediately pointed this out to those present, saying that these data disproved all that was being said in support of the restrictions upon cyclists' use of roadways and the promotion of bikeways, while supporting my analyses showing that cyclists ought to act like other drivers of vehicles. Thereupon, that study was suppressed and no further copies were made available. I still have my copy.

Then California established the California Bicycle Facilities Committee to prepare other standards for bikeways. During the operation of that committee, the official representative of the Legislature informed the committee that, despite the opposition to bikeways that was occurring in the committee, the Legislature insisted that there be bikeways. In other words, the existence of bikeways was not determined on the basis of the safety or the convenience of the traveling public, but of legislative fiat. Also, during the operation of the committee, I repeatedly said that the committee ought to be working on changes in design that would actually reduce the car-bike collisions that we knew, on the basis of Cross's statistics, were actually occurring. This the committee refused to do, presumably because its members knew that if they did those things they would be undercutting the arguments for bikeways and demonstrating that bikeways were more likely to increase than to decrease accidents.

So far as I could tell, the purpose of the committee was to produce designs for facilities that would get cyclists off the roadway for the convenience of motorists, and little happened that had any other import. The most effective work of cyclists on that committee was to prevent the adoption of the most dangerous proposals, and we accomplished that by presenting credible scientific bases for suits against the government by injured cyclists. The net result of the work of that committee was a set of standards that got cyclists off the roadways, that had no bearing on reducing accidents to cyclists, but at least did not intensely endanger cyclists.

At the same time, the federal government had published its set of standards, Safety and Location Criteria for Bicycle Facilities. I have a copy; I don't know whether you do. I criticized these also for their dangers to cyclists and for their scientific errors. Although published, these were never adopted. Because the California standards were less likely to involve the governments in liability for injuries to cyclists, the states and the federal government adopted the California standards.

By this time we had more knowledge about accidents to cyclists. Cross's second study, [2] of a nationally representative sample of car-bike collisions, showed that bikeways were directed at only 2% of car-bike collisions, and other studies of different kinds of cyclists showed that these constituted only 0.2% to 0.3% of accidents to cyclists. In other words, we have allowed only 0.2% to 0.3% of accidents to cyclists to determine the national cycling policy. On scientific grounds that is utterly absurd; as far as cyclists are concerned it is unjustifiable discrimination.

I have had two purposes in telling you this history. The first, and less important, is to inform you of my part in the history and theory of bikeway standards and of cycling theory. The second, more important reason, is to give you the understanding that the AASHTO standards were never intended to reduce accidents to cyclists or to make cycling safe for unskilled cyclists. These considerations were never a part of the deliberations of the committee that produced them, (nor have they been a significant consideration of those who have later revised them). Although ill-informed people have made these claims, nobody has ever made a reasonable demonstration of either of these claims. The only times that safety was considered was in attempts by cyclists to reduce the added dangers to the level of normal cycling on normal roads, and we did not completely succeed in that. Bikeways per the AASHTO standards are more dangerous than normal roads, although they are far safer than the initial standards.

3 Cyclists Need to Act Like Drivers of Vehicles

The second important principle is that cyclists need to act and be treated as drivers of vehicles. This is the only option that is safe and practical.[3]Not acting as a driver of a vehicle, as is often encouraged by bikeways, causes the cyclist to ride into danger unless he slows and accepts delays to prevent those new dangers from injuring him. Nobody knows how to design a ground-level cycling system in which the cyclist acts in some other way than as a driver of a vehicle, without putting the cyclist to increased danger and, therefore, to increased delay to prevent the increased danger from injuring him.

4 The Skill to Act Like a Driver Is Easy To Learn

The third important principle needed to understand the bikeway controversy is that it is not difficult for cyclists to act as drivers of vehicles. Today, young cyclists largely cause their own car-bike collisions by disobeying the principles of traffic operation that underlie the traffic laws. [4]However, it has been demonstrated [5] that third-grade children can learn, in 15 class hours of instruction, to ride better than average adults on two-lane residential streets. Older children can learn, in the same time, to ride far better than average adults on any kind of street at all. The ability to cycle properly and lawfully is present in almost all people from a quite early age. Contrary to fact, bikeway advocates claim that cycling aptitude is rare and actual cycling ability is rarer still.

5 Cycling Skill Reduces Accidents by a Large Factor

The third important principle is that the skill of cycling properly reduces accidents by about 80%, as shown by comparing several studies [6]of accidents to cyclists. That is, the training described above and the additional understanding that that behavior later develops is likely to reduce accidents by that much. Contrary to fact, bikeway advocates claim that only changes to facilities can reduce accidents to any significant extent, and that bikeways do so.

6 Opinions About Traffic Are Determined by Fear and Superstition

It is well known that most people, including traffic engineers, characterize the dangers to cyclists as according to the character of the same-direction motor traffic. It is equally well known, in a scientific sense, that this traffic causes only a very small portion of accidents to cyclists. This paradox demonstrates that the opinions of most people about the dangers to cyclists are incorrect. This false opinion has been created by the bike-safety programs that for decades have taught people that they will be killed if they get in the way of cars.

This false opinion has two effects. First, it causes people to ride dangerously because it directs all their attention and actions to prevent being hit from behind, and therefore prevents them from paying attention and taking action about the actual dangers, nearly all of which are visible in front of them. Second, it provides the public justification for bikeways, whose only safety function is to prevent being hit from behind. The fact that this opinion has been inculcated from childhood through fear of death accounts for the absurdity that the public demands bikeways that cannot make cycling significantly safer, and in practice make it more dangerous.

For further information on this subject refer to my works Effective Cycling and Bicycle Transportation.

7 Defects of the Design-Cyclist Approach in AASHTO

The design-cyclist approach comes from a document, Selecting Roadway Design Treatments to Accommodate Bicycles, written by William Wilkinson for the FHWA.[7]

7.1 Errors in Scientific Procedure

This document has been the subject of considerable correspondence between me, John Fegan, Wilkinson's project manager in FHWA, and Congressman Tom Campbell, which is attached. The important procedural facts are described in my letter of 4 September 1991 to Congressman Campbell. The research project was not on the initial list of proposed projects circulated for review. Wilkinson and Fegan disagree on how the scope of the project was determined. Fegan initially required safety to be the main criterion, but Wilkinson added encouragement, apparently on his own initiative. Wilkinson published his conclusions about 1 September, 1991, after which Fegan told me that Wilkinson's research plan had not been submitted and was not due until 15 September. I have never been sent any research paper on which Wilkinson's report was supposed to be based. In short, the conclusions of this paper were pre-determined and directed toward encouragement instead of safety. No actual physical research was done, so far as I have been able to determine

7.2 Scientific Errors:

Erroneous Classification of Cyclists

Both Wilkinson and AASHTO use the same three classes of cyclists: A (Advanced), B (Basic), and C (Children), described in roughly the same concepts.

Wilkinson describes Class A cyclists as "These are experienced riders who can operate under most traffic conditions. They comprise the majority of the current users of collector and arterial streets." AASHTO describes these as "Experienced ... commuting/utility and touring."

Wilkinson describes Class B cyclists as: "These are the casual or new adult and teenage riders who are less confident of their ability to operate in traffic without special provisions for bicycles. Some will develop greater skills and progress to the advanced level, but there will always be many millions of basic bicyclists." AASHTO uses these words: "casual, novice, occasional, recreational."

Wilkinson describes Class C cyclists as: "These are pre-teen riders whose roadway use is initially monitored by parents. Eventually they are accorded independent access to the system. They and their parents prefer the following:" AASHTO simply describes them as "pre-teen."

Both Wilkinson and AASHTO say that it is appropriate to combine Classes B and C for most purposes and recommend that this standard of an untrained, incompetent, immature, less than twelve-year-old cyclist determine the design of most of the bicycle facilities in the United States.

Wilkinson estimates that fewer than 5% of American cyclists qualify as "experienced or highly skilled bicyclists," Class A cyclists. AASHTO says that "Studies have shown that a very low percentage of current bicycle owners qualify as experienced or highly skilled bicyclists." Actually, there have been no such studies. What constitutes an experienced cyclist? The concept is deliberately ambiguous. We probably consider a motorist with 5 years of experience to be experienced and we certainly let motorists drive on arterial streets after only 30 days of experience. Most cyclists by their teen-age years have had 5 years of cycling experience and therefore qualify as experienced cyclists, by definition. What constitutes a highly skilled cyclist? Wilkinson's criterion is one who rides on arterial roads. AASHTO's criterion is one who participates in commuting, utility cycling, or touring.

Wilkinson's study was criticized on the grounds that if we produced a useful amount of cycling transportation, nearly all of those who do it must be experienced, as it is impossible to have an ongoing transportation system in which most of the users are inexperienced in it. Wilkinson's theory is as silly as saying that most of the married population are inexperienced virgins. The proportion of experienced users must be more like 95% than the 5% stated by Wilkinson. The result of Wilkinson's theory was that he recommended designs that he said were suitable for inexperienced B/C cyclists on nearly all streets except those that carried high-volume, high-speed traffic. That confirms that this document intends that our cycling transportation system is to be designed for use by children.

AASHTO describes Class A cyclists by their purposes in cycling: commuting cycling, utility cycling, and touring cycling. By AASHTO's own definition, people who cycle for these purposes are experienced. In any cycling transportation system, these must constitute the majority, but AASHTO's policy is still based on its statement that only "a very low percentage of current bicycle owners qualify as experienced." That is what is known as a weasel-worded statement. Certainly, few among today's (current) bicycle owners use their bicycles for commuting or other transportational purposes, but any transportation system must be designed for those who use it, not for those who do not yet use it.

The same analysis applies to the consideration of level of skill. Wilkinson and AASHTO define highly skilled only as being able to ride on arterial streets, without describing that skill in any other way. They work on the assumption that some magical effect protects cyclists on arterial streets. Actually, being able to cycle safely on arterial streets is only the basic level of competence for any teen or adult cyclist. Forester taught thirteen-year-old children to do so better than average adults in 15 class hours, and the Palo Alto schools continued that program for several years. All that is necessary is to know how to obey the basic principles of traffic. Anybody who does not know this is unsafe in any urban traffic except that on the easiest residential streets.

Finally, the classification scheme predicts the continuation of the status quo ante, although the AASHTO Guide is based on changing present conditions to a very large extent.

In summary, the classification scheme is not valid because it is based on inherent self-contradictions that cannot be corrected.

7.3 Scientific Errors:

Safety Issues

Both Wilkinson and AASHTO correctly write that the different classes of cyclist have different preferences in bicycling facilities. Both of them then assert that all of these cyclists are best served by providing the facilities that the majority of their class prefer. That transition from uninformed preference to best service is unjustified. Transportation has objective criteria of allowing people to move conveniently, efficiently, and safely. People who demand the ability to move in ways that endanger themselves should not be encouraged to do so. Demands that society provide dangerous facilities for use by all should not be accommodated. Those who unwittingly demand dangerous facilities in the belief that these facilities will protect them should be told the truth, their demands should not be met, and they should be directed toward safe practices instead.

The assertion that bikeways make cycling significantly safer than following the standard highway designs and traffic laws must be demonstrated by those who make that assertion. Despite decades of demonstration bikeways and scientific projects directed to demonstrating that hypothesis, it has never been demonstrated. I repeat, no study has ever demonstrated that bike-lane stripes, bike paths, or bike routes make cycling significantly safer. [8]

The assertion that bikeways are dangerous also needs to be demonstrated by those who make the claim. For bike paths, the evidence is strong.

First, Kaplan's study of members of bicycle clubs showed that the accident rate per bike-mile on bike paths was 2.6 times that on normal roadways, and that car-bike collisions are only marginally more serious than are other types of accidents to cyclists.

Second, the European nations with extensive bike-path systems have only recently started to study the safety of those systems. They have concluded that bike-path systems reduce car-bike collisions between intersections but increase them at intersections. Since most car-bike collisions occur at intersections and driveways, increasing the majority causes an increased total number of collisions.

Third, it should be obvious that most bike paths in the typical urban area must be installed alongside existing roadways and, for any given trip, must cross the same amount of traffic that the roads do. It is impossible that such paths can avoid being more dangerous than well-designed roads. Cyclists have found that such sidepaths, when attempted to be used at normal road speeds, are the most dangerous facilities that we have. Even when used by untrained, slow cyclists, predominantly children, such sidepaths in Palo Alto have shown a car-bike collision rate 1.8 times that on the road. [9]

Fourth, analysis of conflicting movements at intersections that include bike paths show that bike paths create more conflicts than the normal designs and that these conflicts are much more difficult for the drivers to handle. The only way that grade-level intersections that include bike paths can be made safe is to install special phases in the traffic signals, thus reducing the green time available to each class of user and therefore creating delays for all users. These traffic signal phases don't reduce accidents that would otherwise have occurred; they only reduce the accidents that have been created by the design.

The old Guide at least admitted that bicycle sidepaths have been found less safe than adjacent roadways. The new Guide specifically deletes that statement at draft pg 19, line 2. The change is a falsehood.

The only paths that have the potential of reducing accidents are those alongside physical barriers to traffic, such as waterways, and even then the potential is rarely realized.

For bike lanes the case is less obvious but still convincing. The potential for reducing accidents is very small, the 0.3% or so of urban accidents to cyclists (2% of car-bike collisions) that are caused by motorists overtaking lawful cyclists. However, the potential for increasing accidents is much larger. This is because bike lane stripes keep cyclists inside the bike lane when they should be outside it, and keep motorists outside the bike lane when they should be inside. [10] The efforts to correct this situation by dashing the bike-lane stripe do not work and cannot work. Drivers (motorists and cyclists) do not understand the symbolism and, in any case, the dashed line is far too short in most places where it exists and it does not exist in most places where it is required. Furthermore, there are many places where nobody has been able to design a bike lane that does not create more conflicts than previously existed, as the long deliberations of the California Bicycle Advisory Committee on this subject have demonstrated. The conflicts created by bike lanes are those that, on normal streets, cause 30% of car-bike collisions. [11] Any significant increase in this 30% is going to outweigh even the maximum possible decrease in the 2%. The only study of accident rates on American bike lanes that appears in the record is of two bicycle lanes in Madison, WI. While the authors tried not to admit it, the data of that study showed that the accident rate per bike-mile increased somewhat.[12]

7.4 Scientific Errors:

Issues of Cycling Skill

In claiming that bikeways provide the best service for beginning cyclists, although not for competent cyclists, the Guide is claiming that bikeways reduce the level of skill that is required for safe cycling. This is not a new claim. As with the claim of reduced accidents, it must be demonstrated by those who make it. No study supporting the hypothesis that bikeways reduce the skill level required for safe cycling has been published over the twenty years that this claim has been made. For that matter, none of those who have made this claim have ever identified the particular skill that cyclists using bikeways can, they assert, safely do without. Without identifying that skill, the claim is nonsense.

However, one study has been made of the behavior and skill that is required for safe cycling on roadways and on bikeways, [13] and one study has been made of the actual behavior of cyclists (and of motorists, to some extent) with regard to bike-lane systems. [14] The study of the behavior and skill that is required shows that cycling on normal roadways requires only the same skills that all drivers need, [15] but that cycling where bikeways interact with roadways requires abilities of perception greater than humans possess. Furthermore, the dangers are not obvious unless the cyclist already understands the normal traffic pattern.

In other words, safe cycling around bikeways is not possible unless the cyclist already possesses the skill of cycling safely on normal roadways. This is directly contrary to the claim made by the Guide that bikeways decrease the required level of skill.

7.5 Scientific Errors:

Planning Issues

The Guide describes best service for B and C cyclists as "providing designated bicycle facilities (bike lanes, separate bike paths, or side street [bicycle] routes) through key travel corridors typically served by arterial and collector streets." Any useful bicycle transportation system designed according to this instruction will be far different from that implied by its words. Nearly all separate bike paths through key travel corridors will be bicycle sidepaths; there are few other places for them in the typical urban area. These are, they cannot be anything else, the extremely dangerous bicycle sidewalks that the Guide considers Class III bikeways and specifically disapproves. The opportunities for what the Guide considers adequate bike paths are so limited that they can be built only in very few locations and cannot form a system.

Bicycle routes on side streets are possible through residential areas, but are generally impossible through commercial or industrial areas. In residential areas they serve slow-speed, short-distance cyclists, but can hardly form a transportation system and, in any case, cyclists with these limited performance desires are quite adequately served by existing residential streets.

That leaves bike lanes. In his document, Wilkinson admits that his plan requires bike lanes on all major streets, and gives specific recommendations according to traffic volume and speed. The Guide pretends that some other system can be produced, but the authors make no effort to show that this is possible. The Guide is mere wishful thinking in this matter. Any useful transportation system produced according to the Guide will consist largely of bike lanes along almost all arterials and collector streets.

7.6 Scientific Errors:

Issues of Transportation Efficiency and Convenience

One criterion for transportation systems is convenience. Inconvenient systems produce waste and reduce opportunities for those who waste time using them or, equally, fail to benefit from useful transportation because the system is not convenient. Inconvenient systems cannot be said to provide good service. While bicycle routes on side streets may serve slow-speed, short-distance cyclists, who are child cyclists and some adult cyclists with only local transportational needs, any system that allows cyclists to generally move about the city must allow travel on all arterial and collector streets. Since the Guide says that only Class A cyclists use these streets without special bicycle facilities, to produce a system that is useful for Class B cyclists requires bike lanes on all arterial and collector streets.

8 The Ethical Question of Providing Best Service: Enticement Of Unskilled Persons Into Activity For Which They Are Not Qualified

If implemented, the Guide must result in a system of bike lanes along arterial and collector streets. The Guide asserts that Beginning and Child cyclists will be best served by riding in bike lanes on arterial and collector streets. The bike lanes do not reduce accidents to cyclists but probably increase them, as demonstrated above. The bike lanes do not lower the level of skill required, but require more skills, as demonstrated above. The Guide states that installation of bike lanes should be done "where it is the goal to increase bicycle use," and they "can be used to accommodate Group B/C bicyclists and encourage potential bicyclists." In other words, the bike lanes are not installed for the benefit of experienced cyclists, but to entice beginning and child cyclists, who would not otherwise use such streets, onto arterial and collector streets. That is the guiding motive of the bicycle plan implied by the AASHTO Guide.

By AASHTO's own definition, these people are not qualified to ride safely on such streets, whether or not such streets have bike lanes. Yet the AASHTO Guide is based on the knowledge that beginning and child cyclists will be enticed onto these streets by the belief that the bike lanes protect them from accidents and allow them to ride safely without knowing how to do it. The same conclusion applies to most urban bike paths. No other conclusion is rationally possible about the motivation and purpose of the AASHTO Guide, insofar as those apply to the cyclists who are presumed to be its prime beneficiaries. It is possible, of course, that the AASHTO Guide serves other purposes that have little relevance for cyclists, or was written by people who do not understand what they have done.

Whichever combination of the above conclusions is correct, the AASHTO Guide presents a completely unethical program of endangering cyclists, either to encourage them or for otherwise undisclosed reasons.

9 Political and Social Reasons Why the Guide Is Bad for Competent Cyclists

The above discussion shows how the Guide is dangerous and unethical for beginning and child cyclists. On first consideration, it would appear that competent cyclists should not worry because they have the skills to avoid the dangers presented by the Guide. That is incorrect. In many cases competent cyclists must operate by obeying the dangerous system. Laws require cyclists to use bike lanes, and even bike paths in some states, wherever they exist. Most urban bike paths cannot be made safe. In some locations (for example, many freeway overpasses) no design of bike lane will describe safe operation for all users, but there is no guarantee that bike lanes will not continue to be installed in such locations.

Furthermore, there are other reasons than safety why the Guide is bad for competent cyclists. The Guide is society's proclamation that normal streets with significant traffic are not safe for cyclists. That means that those cyclists who choose to ride on such streets are risk-taking fools who do not deserve the respect or protection of society. Even where the laws do not require cyclists to use bikeways of either type, those motorists who believe that cyclists should use bikeways instead of normal roadways are encouraged by the existence of the Guide and by the bikeways produced by the Guide to harass cyclists who use normal roadways. Even those people who do not openly harass cyclists are influenced by the existence of bikeway programs to assume that cyclists are, as the Guide says that they are, immature and incompetent, and are thereby of lower social standing than normal people.

10 It Is Absurd To Design a Transportation System For Use By Beginners

The Guide specifies a cycling transportation system that is designed for use by beginners. The highway system is an enormous investment that is designed for use by people who know how to use it. That is the correct approach for two reasons. First, most users cannot be beginners. Second, designing the system for beginners increases costs and decreases efficiency.

11 The Motives For The Guide Remain Suspect

The initial motive for standards for bikeways was to get cyclists off the roadways for the convenience of motorists. This was never openly admitted, but no other explanation that the promoters offered either met the facts or survived the test of reason. Of course, the promoters believed that cyclists delayed motorists and that delayed motorists hit cyclists, so that delay and danger became synonymous in their minds, but that convenient superstition was contradicted by the facts that were known at the time and have been reinforced by further studies since then. The insufficiency of the arguments on which the Guide is based, indeed their immaturity, shows that little has changed. Therefore, the most reasonable explanation for the Guide remains that it is intended to get cyclists off the roadways for the convenience of motorists.

12 The Guide Is A Discouragement To Cycling Transportation

The Guide openly says that the bikeway program that it specifies is intended to encourage cycling transportation. That is correct in only one sense. The Guide does not encourage cycling transportation by competent people, it encourages only cycling transportation by immature and incompetent people. Bicycle advocates who make excuses for bikeway programs express the hope that these immature and incompetent people will take up cycling to such an extent that they will, in time, become competent. In contrast, the Guide assumes that this will not occur, that those people who are attracted to cycling transportation will largely retain the characteristics of those that it considers beginners. This is a very real possibility, but it is not the only problem with this prediction.

Despite two decades of the argument, no study exists showing that bikeways have attracted many people to cycling transportation. Two kinds of studies exist, those correlating cycling transportation to existing conditions and those investigating the opinions of non-cyclists. The correlation studies show more cycling where there are bikeways, but also show greater correlation with other factors, such as the presence of large university cycling populations, in the U.S. [16], and of enormously inconvenient motoring conditions, as in Amsterdam. [17] In other words, cycling transportation tends to exist where it provides objective advantages, and the presence of bikeways is more a response to that cycling transportation than the cause of it. The opinion studies [18] show that non-cyclists believe that bikeways make cycling safe. However, these studies do not show, and their method cannot determine, that providing bikeways will cause these people to cycle for transportation. They may choose not to do so for a variety of reasons that cyclists recognize are very important in making the decision to cycle for a particular trip. The fact that no study yet shows that bikeways create cycling transportation demonstrates that the connection between non-cyclists' opinions about bikeways and cycling transportation cannot be more than extremely weak, and probably does not exist at all.

Not only does the bikeway program specified by the Guide fail to create cycling transportation, its emphasis on designing for, and therefore maintaining, the cycling style of beginners is an active discouragement to cycling transportation. The physical arrangement of the bikeways itself makes cycling less convenient and the political effects denigrate cyclists, but there is a further educational effect. The opinion polls mentioned above clearly show that the public believes that bikeways make cycling safe. We know that that superstition is false, but the Guide's bikeway program proclaims the truth of that superstition. People who believe that bikeways make cycling safe do not make the effort to learn how to ride safely. [19]

The effect is much more marked in the nations with extensive bike-lane systems, those in northern continental Europe. European cyclists used to be considered accomplished, the epitome of competence, able to ride safely in normal motor traffic. The skills that competent cyclists practice and advocate here were first learned from European cyclists. However, that has changed in those nations with extensive bike-lane systems. Cyclists there, and their traffic engineers for that matter, have no idea of how to ride properly. They ride dangerously, as their bikeways require, they suffer the consequences of that dangerous cycling, and only now, after many decades of this maltreatment, are they beginning to realize that it is their designs that produce the excessive car-bike collisions. They have not yet recognized that their designs also have produced a cycling-incompetent population, but that is obvious to any competent cyclist who observes their urban cycling.

12.1 European Conditions Differ Greatly From Those In The U.S.A.

There are those who recommend the Amsterdam style of cycling transportation, but that will not work in the U.S.A. It works in Amsterdam because there motoring is very inconvenient and distances are short. However, the volume of cycling transportation in the Netherlands is dropping, has been dropping for decades, while the volume of motoring is increasing, as Dutch conditions become more like those here. Dutch officials have been quoted (1995) as saying that their program has failed and that they need to reconsider what should be done. Modern American cities, with their decentralized design and long distances, require that cyclists ride fast in order to cover the distances in acceptable time, which requires cycling on well-designed roads with the rights and duties of drivers of vehicles, not on typical bikeways.

13 Conclusions Regarding the Classification of Cyclists in the AASHTO Guide

Designing bicycle transportation systems according to the classification of cyclists specified by the AASHTO Guide has several detrimental effects:

1:It does not reduce the accident rate and probably increases it.
2:It entices untrained people into an activity that requires skill for safe participation.
3:It does not reduce the level of skill that is required, but rather increases it.
4:It greatly reduces the incentive to learn safe cycling skill.
5:It denigrates competent cyclists, lowering their status in society and before the law.
6:It produces a system that is more suited to Dutch cycling conditions than to American cycling conditions.
7:It does not encourage cycling transportation, but rather limits its appeal to that done by beginners.

14 A Classification System and Cycling Transportation Program Without These Defects

14.1 Wide Curb Lanes

For General Transportation

Because any successful urban highway transportation system will be used primarily by experienced users, it should be designed for experienced users. By the Guide's own statements, the best facility for bicycle transportation by experienced users is the wide curb lane. However, the recommendation for wide curb lanes is not based on a verbal rationalization of the Guide's words, but upon engineering analysis with sound common sense.

The wide curb lane provides space for the motorist to overtake the cyclist without using the adjacent lane, and it does this without any of the adverse effects of bike-lane stripes. By doing so it answers the major concern of motorists and highway organizations about bicycle traffic, that bicycle traffic will delay motorists and cause congestion. Fear of such delay is the deduced reason why motoring and highway organizations have promoted bikeways, since no other explanation suffices. Wide curb lanes answer that concern.

Wide curb lanes also have the following advantages over bike-lane stripes:

1:They do not confuse motorists and cyclists about how to drive properly, causing them to make dangerous mistakes.
2:They do not proclaim that it is safe to cycle without the required skill.
3:They do not entice inexperienced people into situations that require competence for safe operation.
4:They do not discourage people from learning proper and lawful cycling technique.
5:They do not encourage the common superstition that normal roads are too dangerous for cycling and that cyclists should not be using normal roads.

The above reasons indicate why wide curb lanes are safer than striped bike lanes. It may be argued that bike-lane stripes reduce motorist-overtaking-cyclist car-bike collisions. As discussed above, these are only 2% of urban car-bike collisions, and there is no evidence whatever that bike-lane stripes are more effective than wide curb lanes in reducing this small proportion of car-bike collisions.

Wide curb lanes are not needed everywhere. They are desirable on those roads that provide useful transportation corridors and attract significant volumes of motor traffic. That is, on arterial streets and some collector streets. They are not needed on streets where the motor traffic volume is low, because the probability of motorist delay is small and the probability of motorist-overtaking-cyclist car-bike collision is very small.

14.2 Special Facilities for Children

Without a doubt young children cannot safely cycle under all the conditions in which adults can safely cycle. However, young children have a smaller need for transportation because their parents let them go to a smaller number of places, whether or not they go by bicycle. Children of eight years of age can safely ride on two-lane residential streets. This is about the youngest age at which children ought to be allowed to cycle without supervision. Children of thirteen years of age can safely ride on multi-lane streets with typical traffic. [20] In each case, the children can ride much better than the average adult cyclists using those same streets. Between these two ages there is a valid need for bicycle facilities that serve this population by providing facilities that enable them to reach suitable particular places without exceeding their expected level of skill. Such facilities would be desirable wherever the desired destinations cannot be served by two-lane residential streets. Bicycle lanes do not allow cyclists with this level of skill to ride safely. Only bicycle paths that are used slowly and have very strong protection of right-of-way at intersections, as with traffic signals with specific signal phases for cyclists, meet this criterion. The difficulties of installing such facilities limit the number of locations where this service can be provided. The inconvenience that such facilities impose on adult cyclists prevents them from being satisfactory for most transportation needs of adult cyclists; therefore, there must not be any implication that adult cyclists should use them. It is entirely satisfactory for those adult cyclists who are accompanying child cyclists to use these facilities; therefore, there should be no prohibition against adult cyclists using them at the low speed for which they are safe.

14.3 Transportational Bike Paths

There are some locations where transportational bicycle paths would be suitable. There are two general types of location. The first is the shortcut that connects roads that otherwise have no direct connection. Such shortcut paths are generally short, without intersections except at their ends, and although they may have to be used slowly and with delays while yielding to traffic at one or both ends, they enable the trip to be made quicker. The other type of location is along some topographical feature, such as a waterway, that has cut the street pattern, so that the path can be built with very little cross traffic. As long as any of the few intersections with streets with heavy traffic are provided with either grade separation or signal phases for the cycling traffic alone, such paths can be useful for transportational purposes, if the location naturally generates such traffic. However, experience has shown that such paths often develop the type of traffic that limits their safe use to low speeds only. It may be that the need to travel slowly exists mostly at non-commuting times and days, but such luck should not be counted upon.

14.4 Recreational Paths

Recreational paths are a different matter. These are designed not for transportation but for the pleasure of riding on them. Such paths are not intended to be safer than normal roadways; some paths may be safer than normal roadways while other paths may be more dangerous. Some may be designed for high-speed cycling, others for only low-speed cycling, but in any case it is likely that the type of traffic they develop will limit their use to only low-speed cycling. These should be built in those locations where the pleasure can be obtained, but they should be built with recreation funds rather than transportation funds.

14.5 Education is Required for Any Program

As discussed above, the only type of bikeway that lowers the level of skill required for safe operation is the not-yet-achieved bicycle path that is protected at all intersections by traffic signals that have special signal phases for bicycle traffic. Even on such a path, the users have to know about traffic signals, and it is almost certain that to reach such a path they would have to travel on at least two-lane residential streets.

If the Guide is seriously concerned with the safe operation of a cycling transportation system, as it says that it is, then it must address the issue of training of cyclists in proper operation of the system. This is true even if the Guide pretends that it concerns only facilities, because it pretends that its bikeways make cycling adequately safe for incompetent people. The Guide must stop pretending that its designs of bikeways enable untrained and incompetent users to operate safely. The Guide must openly state that no practical designs of bikeways or highways allow safe operation of a cycling transportation system by untrained, incompetent persons. The Guide must openly state that its designs can only be used safely by people who know how to ride properly on normal roads. It goes without saying more here that the Guide must also eliminate all of its designs that require a higher level of skill, as discussed above.

Since the Guide discusses other subjects than merely the design of bicycle facilities, it is highly desirable that the Guide provide some guidance about the level of skill that is required to operate safely on its designs. I offer the following five traffic-cycling principles that all adult (13 years and up) cyclists ought to be able to follow.

1:Ride on the right-hand side of the road, not on the left and never on the sidewalk.
2:When approaching a road that is larger than the one you are on, or has more or faster traffic, you must yield to traffic on that road. Yielding means looking and waiting until you see that no traffic is coming.
3:When preparing to move laterally on a roadway, you must yield to traffic in that line of travel. Yielding means looking forward and backward to see that no traffic is in that line of travel.
4:When approaching an intersection, you must choose your position according to your destination. Right-turning drivers are at the curb, left turning drivers are at the center, while straight-through drivers are between them.
5:Between intersections, you choose your position according to your speed relative to other traffic. Parked ones are at the curb, medium-speed drivers are next to them, while fastest drivers are near the center of the road.

To operate on multi-lane streets requires being able to operate in accordance with all five of these principles. Ten-year-old children can learn how to do this. To operate on two-lane residential streets requires being able to operate according to the first three of these principles. Eight-year-old children can learn how to do this.

15 Conclusion

The only known cycling transportation system that can be safe and useful under American conditions is one in which good standard roadway design, including wide curb lanes, is used by cyclists who know safe and lawful cycling technique. This may be supplemented by very carefully designed bicycle paths in the few locations where paths provide clear advantages. The AASHTO Guide for the Development of Bicycle Facilities ought to specify such a system.


1 The first general publication was: Forester, John; Cycling Traffic Engineering Handbook; 1977. The current presentation of this analysis is in: Forester, John; Bicycle Transportation, 2nd ed.; The MIT Press, 1994.

2 Cross, Kenneth D. & Gary Fisher; A Study of Bicycle-Motor Vehicle Accidents: Identification of Problem Types and Countermeasure Approaches; NHTSA 1977

3 The correct method of cycling, and how to learn it, are described in: Forester, John; Effective Cycling; 1976; 6th ed., The MIT Press, 1993.

4 The age distribution of types of car-bike collision is given in Cross's data and is well charted in Forester's Bicycle Transportation.

5 Forester, John; Elementary-Level Cyclist Training Program: Objectives, Techniques and Results; 1981. Forester, John & Diana Lewiston; Intermediate-Level Cyclist Training Program: Objectives, Techniques and Results; 1981

6 Kaplan, Jerrold A.; Characteristics of the Regular Adult Bicycle User; Master's Thesis, U of Maryland; NTIS, Springfield, VA; 1976

Chlapecka, T.W., S.A. Schupack, etc.; Bicycle Accidents and Usage Among Elementary-School Children in the United States; Chicago; National Safety Council; 1975

Schupack, S. A. etc. Bicycle Accidents and Usage Among Young Adults; Preliminary Study; National Safety Council; Chicago, 1976
Watkins, S. M., Cycling Accidents Among Members of the Cyclists' Touring Club; CTC, England, 1984

7 Wilkinson, W.C. III, A. Clarke, B. Epperson, R. Knoblauch; Selecting Roadway Design Treatments to Accommodate Bicycles; FHWA-RD-92-073, Jan 1994

8 Reduction of car-bike collisions by installation of a bike lane was claimed in San Diego. However, there the accident reduction was produced by prohibiting parking of the large vehicles that typically were parked along that street. This reduced ride-out accidents among young children. Prohibiting parking does not depend on a bike lane.

9 Wachtel, Alan, Diana Lewiston, Gayle Likens; Risk Factors for Bicycle-Motor Vehicle Collisions at Intersections; ITE Journal, Sept 1994.

10 Forester, John; Effects of Bike-Lane System Design on Cyclists' Traffic Errors; 1978

11 Data from the Cross second study, analyzed in Forester's Bicycle Transportation

12 Smith, R. L. Jr., & T. Walsh; Safety Impacts of Bicycle Lanes; TRR #1168, pg 49-56, 1988

13 In Forester's Bicycle Transportation.

14 Forester's bike-lane study.

15 Motorcycle Driving Task Analysis; Motorcycle Safety Foundation

16 Goldsmith, Stewart; Reasons Why Bicycling and Walking Are Not Being Used More Extensively As Travel Modes; National Bicycling and Walking Study #1; FHWA-PD-92-041.

17 No formal studies, but see Forester: Bicycle Transportation for this obvious analysis.

18 There are so many that it is unnecessary to specifically list any one, considering the conclusion.

19 Forester's study of bike-lane systems shows that students of UC Berkeley, without a bike-lane system, have learned to ride properly, while students at UC Davis, with a bike-lane system, have not learned to ride properly.

20 Forester's studies of elementary-school cyclists.

Return to: John Forester's Home Page                                                 Up: Government