Bicycle-Safety-Related Research Synthesis:

written by Andy Clarke and Linda Tracy of the Bicycle Federation of America

Reviewed by John Forester

home             government

1 Introduction

1.1 The Importance of this Document

The Synthesis of Bicycle Safety-Related Research is the supposed scientific justification for the cycling policy and program of the nation. It purports to contain a summary of all the important research since 1981 into cycling transportation. That information, when added to the valid portions of the work done before 1981, ought to support our nation's cycling policy and program. According to Clarke and Tracy, it does. (Hereafter, I list the authors as C&T.) However, they have made it do so only by ignoring half or more of what was done and by misquoting and maligning much of the rest. The result probably persuades the politicians and highway administrators, and many bicycle advocates as well. However, to one who knows what actually was done and the amount of scientific support possessed by each report, the result demonstrates that our nation's cycling policy and program are so far from reasonable that, instead of being supported by scientific knowledge, they can be supported only by ignorance and/or mendacity. That demonstration is the important work of this document.

1.2 Scientific Work Before 1981

The work that had been done before 1981 was the basic research into cycling accidents and into cycling and driving methods, and the creation of the standards for bicycle facilities (first California's; then that became the national one) that should have been based on such research. However, the work was not done in rational sequence. The standards for facilities were started first, and the first ones were rejected. The dangers of these to cyclists elicited the first research into cycling and driving methods, done by cyclists, which cyclists used to convince those who produced the standards to retract the most dangerous parts of them and to demonstrate that cyclists fare best when they act and are treated as drivers of vehicles, if only the designers would let them. Finally, after the standards were complete, came the results of the research into cycling accidents, which should have been the basis for the initial work, but which demonstrated that the standards were more likely to increase car-bike collisions and other accidents than to decrease them.

To argue that the facility standards were intended to make cycling safe and indeed do so requires either ignoring or maligning all the other work of the period. C&T ignore the cycling traffic engineering work that cyclists did during the creation of the standards. While C&T list some other accident statistics without comment, they argue from only one data point of Cross's statistics on car-bike collisions, and misapply even that one. C&T misquote and then malign Kaplan's study. C&T could hardly ignore the opposition that was known to exist to the government's policy and program, but they misplaced this to the next decade and blamed it all on the work of teaching cyclists to ride safely, which they derided as unpopular and elitist.

1.3 What the Authors Say About Their Document

The authors state that this is not a synthesis of later bicycle-safety-related research, giving the excuse that little such research has been done since the 1970s. Instead, the authors describe their view of the cycling program policy that their employers and financial sponsors, the Bicycle Federation of America and the Federal Highway Administration, happen to prefer. The authors also quote documents produced or published by their employer, William Wilkinson and the Bicycle Federation of America, without disclosing these connections.

1.4 The Authors' Method

Consideration of what was included and what was excluded suggests a concerted plan to mislead the public, to put a false front on a policy and program that C&T know to be false. That is, knowing the policy and program to be false, C&T made up a lists of what to include, what to exclude, and what to malign, based on how well each item supported their argument. If that is so, they made a pretty good job of it. An argument that was produced merely from a smattering of knowledge in a field of ignorance would not have been planned so well; the authors would have found themselves caught up in contradictions. However, there is one more argument beyond those of mendacity and ignorance. That is the argument of overriding superstition. The authors may believe that they made an honest job of writing this study and by their beliefs they did so. The problem is that they may so believe the superstition that same-direction motor traffic is the prime danger to cyclists, and have so studied the cycling world (most of whose inhabitants believe the same superstition) with that superstition in their minds, that their minds now reject information to the contrary. They may know of its existence, but their minds give it no significance. The only items of knowledge that they pay attention to are those that agree with their superstition. Most Americans, voters and elected officials, believe that superstition, and our cycling policy and program are based on it. It is no wonder then that when government wants to marshal support for its misguided policy, it selects more people who, either pretending to be honest experts or so beguiled by their superstition that they do not recognize their ignorance, believe that superstition. I think that this is the better explanation.

2 Outline of Clarke & Tracy's Statements

Throughout this section, my descriptions of and quotations from the document are in plain type, my comments on those descriptions and quotations are in italic type and are enclosed by braces { }. These are to enable readers with different electronic systems to distinguish the two. The paragraph sequence is that of C&T's document, its chapter 1 being my 2.1, etc.

2.1 Bicycling in the United States in the 1990s


2.1.1 The Potential of Bicycling in the United States

The study quotes the well-known findings that cycling is the only US transportation mode which many more would like to use than actually do use.

C&T assert that "US transportation consumers do not have the same choice of modes available to them as do those in the Netherlands, Denmark, Japan, Australia, and the United Kingdom." {This is false. US transportation consumers have the choice; they merely fail to choose cycling.} C&T make the valid point that one needs a system for cycling transportation, particularly secure parking. They assert that significant cycling will occur only when we have short distances. {This position claims that future American cyclists will not ride the distances that present American cyclists ride. They make designing a system that appeals to non-cyclists their prime directive. Why should we expect that a system that appeals to non-cyclists will be best for cyclists, or that present non-cyclists, once attracted to cycling, will remain non-cyclist in attitude?}

2.2 Bicycle Crash Experience

Quotes many studies of cycling accidents. Nothing special to say, and no judgements rendered except to quote the largest possible number for Cross's same-direction car-bike collision statistics.

2.3 Intersection Countermeasures

C&T praise the Minnesota Bikeway Design Manual as "one of the most complete." For example, that says that one of the causes of intersection conflict is "poor visibility to the right rear of a motor vehicle." {The problem is cyclist overtaking on the right, not the poor visibility to the driver of the motor vehicle. If the cyclist doesn't get into the dangerous position, there is no problem.}

C&T discuss stop signs, but fail to discuss yield signs. They praise Palo Alto's bicycle boulevard for its lack of stop signs. {However, they fail to note that, even so, other routes are still faster for most cyclists who use it. Those who use it do so because they are frightened of same-direction traffic on the faster routes.}

Under traffic signals, C&T discuss methods for making them sensitive to bicycles, and for proper timing. Under right-turn-on-red car-bike collisions, C&T correctly say that many involve wrong-way riders, rather than RTOR itself.

About freeway ramp intersections, C&T make the following absurd assertion: "bicyclists can be channeled to visible crossing points where their exposure to fast-moving motor vehicle traffic is minimized." {Each ramp carries the same number of cars for its entire length. Since the cyclist must cross this line of cars, it is impossible to minimize his exposure. C&T make the cyclist ride down the ramp and then swerve across the traffic, a situation that is so dangerous that the cyclist, to preserve his life, must yield to the traffic.}

About right-turn-only lanes. {C&T fail to recognize the usefulness of right-turn-only lanes in separating right-turning traffic from bicycle traffic. While they complain about the cyclist having motor traffic on both sides of him, they don't recognize that it is far better to have the right-turning traffic on one's right than on one's left. If there are to be right-turn-only lanes they want them short to keep traffic speeds low.} C&T say that short lengths of bike lane on the right side of both straight-through and left-turn-only lanes can be useful.

Advanced stop lines. C&T state that experience has shown that these "can be used satisfactorily on junction approaches with one or two lanes and with vehicle flows of up to 1,000 vehicles per hour." The reasons that they give for adopting advanced stop lines are not persuasive to an effective cyclist: cyclists can then avoid right-turning motor vehicles, can get into position to make a left turn, have time to clear the intersection, and are positioned where motorists can clearly see them. {Advanced stop lines change normal procedures only when the cyclist arrives near the end of the red phase. Advanced stop lines do not reduce the skill required of the cyclist; indeed, they need an additional skill, that of improving his predictions about the state of the signal when he arrives at the intersection. The cyclist still needs to know how to ride properly for all the other times. When the cyclist is approaching the signal, he has little chance of predicting accurately the state of the signal when he actually reaches the intersection. Therefore, he needs to ride in the normal manner during his approach, lest he be caught at the curb when he should have been in the straight-through or the left-turn lane.}

Traffic circles. C&T state that small ones perform a traffic-calming function and reduce accidents to cyclists, while for large ones "no method has been found to reduce the vulnerability of bicyclists in the roundabout." {Of course, the problem is that most cyclists do not use roundabouts correctly; they get caught on the right-hand-side of exiting cars.}

Mid-block car-bike collisions. C&T's discussion confuses bike path intersections with driveways. They recommend the Minnesota warrants for different types of bike path crossings: marked, with island, with traffic signal, grade separated. To prevent motorist mid-block left turns they recommend limiting the number of driveways and installing medians.

2.4 Bicycle Accommodations and Facilities

2.4.1 The Bikeway Debate

C&T introduce their section on bicycle accommodations and facilities with their description of the great bikeway debate. They recognize that the 1970s produced standards for bikeways, without giving any history of how those standards were created or judgement about the type of standards that resulted. Although they do not explicitly state, they apparently approve of these standards and believe that the authors of these standards were motivated by the desire to do good for cyclists and did so.

C&T describe the 1970s as "an intense period of 6 or 7 years of experimentation and research into many aspects of bicycling." {That was so, but in the opposite sense to that they convey. Government was experimenting by imposing on cyclists dangerous designs of bicycle facilities, all for the convenience of motorists. Cyclists were resisting. I killed the first proposed California standard by demonstrating that the designs were dangerous and the first proposed federal standard by demonstrating that the research on which it was based was incompetent.

Out of that controversy, cyclists created the discipline of cycling transportation engineering, first published in my Effective Cycling of 1975 and my Handbook of Cycling Traffic Engineering of 1977, and continued through the sixth edition of Effective Cycling (1) in 1993 and the second edition of Bicycle Transportation (2) in 1994.

We cyclists knew that the proposed designs were dangerous, but initially we could not prove it except by analysis of driving methods. Then the two studies that C&T mention were published: Kaplan's study of club cyclists (6) and Cross's study of car-bike collisions.(7) [There had been an initial study by Cross in California that was suppressed by the California Highway Patrol because it didn't support their goal of getting cyclists off the roads.] These two studies supported the conclusion already reached by cyclists, the vehicular-cycling principle that "cyclists fare best when they act and are treated as drivers of vehicles." Little other cycling research done in that period has stood the scrutiny of time and controversy, but that which has says exactly the opposite of what C&T try to convey.}

C&T then describe the 1980s as a period of little governmental activity but one in which, so they say, the anti-bikeway vehicular-cycling theory was developed and spread by the publication of Effective Cycling. C&T denigrate Effective Cycling by misquoting it and misstating its effect. They say that EC holds that cyclists should "behave like motor vehicles and that bicyclists should not venture onto the roads until they can ride this way." They misstate EC's effect by saying that it has only a "hard core of followers ... experienced and club cyclists" and that only 3,000 cyclists have passed an Effective Cycling course.

Having described the 1970s as the period of producing detailed standards for different types of facility, and the 1980s as a period of regression, C&T then say that the 1990s are the period of selecting which of these types of facility to install in each location.

{Many of C&T's statements are so general that it is difficult to determine their position and discover the errors in it; some will become obvious in later discussion. C&T fail to recognize that the authors of the standards adamantly refused to base the standards on methods of reducing accidents to cyclists; their aim was to get cyclists off the roadways for the convenience of motorists. Therefore, they fail to recognize the opposition of cyclists during the creation of the standards and that the development of the vehicular-cycling principle (as differentiated from the decades-long practice of vehicular cycling) was simultaneous with the development of those standards and was motivated by the disregard, by the authors of the standards, of scientific knowledge about cycling. C&T's date of 1984 for Effective Cycling and their ignorance of Bicycle Transportation show that they do not know the subject.

Effective Cycling says that cyclists should act like people, like drivers and not like machines, and that beginners should "ride on easy traffic roads ... then you can cycle enough to get experience, which will allow you to understand and practice the more advanced habits and maneuvers."

Although C&T mentioned the small number of people who may have taken Effective Cycling courses, they failed to mention the effect of more than 100,000 copies of that and derivative books on the knowledge, skill, and politics of the informed cycling public. Their "hard core" remark is intended to convey that competent cyclists are elitist old fogies who are resisting progress. Far from it; the hard core are the small minority who know cycling. If we were talking about swimming, no-one would give more credence to non-swimmers than to swimmers, yet that is C&T's attitude. Given the amount of information that is available on these subjects, this is, at best, mendacity for a political purpose.}

2.4.2 Facility Selection

C&T write about the 1990s: "For the first time in a decade, there was overt support for bicycling at the national level. More importantly, there was support for increasing use, not just accommodating existing users and making them safer." {This statement shows C&T's ignorance or mendacity about these subjects. It is true that there was a change, but they are inaccurate about both the nature of the change and about the preceding policies. The preceding policies were neither to accommodate existing cyclists, at least not in any sense that refers to cyclists, nor to make cycling safer. The policies under which the standards for bicycle facilities were designed were to get cyclists out of the way of motorists without regard, until cyclists forced the issue, for the safety or convenience of cyclists. Existing cyclists developed the theory of vehicular cycling to protect their right to a safe and convenient cycling transportation system.(3)}

C&T quote William Wilkinson to explain the difference. "Accommodating current use: This policy would focus attention on meeting the needs of only current bicycle users and is consistent with the general approach taken by most bicycle advocates over the past 15 years. These advocates believe bicyclists should have the knowledge and skill needed to operate a bicycle in the traffic conditions typically associated with shared use of streets and highways."

{Wilkinson's description of this as "accommodating current use" is incorrect in at least three ways. First, the system that he considers to be the present, which by his own words is the vehicular-cycling system, can accommodate any possible increase in the number of cyclists. Second, the present system is not the vehicular-cycling system, but that system polluted by different policies for facilities and traffic laws. Its facility policy calls for a bikeway system with the grudging acceptance that prohibiting cycling in the vehicular manner would create a political and liability nightmare. Its legal policy calls for cyclists to act other than as drivers of vehicles, with the grudging acceptance that requiring any other method would be a political and liability nightmare. Third, most current users do not operate as they should, which is one reason for the large number of accidents and for the small amount of cycling transportation done.

Wilkinson is not correct in saying that most bicycle advocates believe in the vehicular-cycling principle. So far as I see, many more bicycle advocates, like himself and the authors of this study, advocate bikeways than prefer the vehicular-cycling policy. By this misstatement, Wilkinson is trying to pass off as old-fashioned, and therefore of no future importance, the minority view that has achieved its importance, not because of numbers, but because it is scientifically correct.

Numbers aside, the very important point is that Wilkinson is trying to make the reader believe that bikeways enable a person to ride for transportation safely without knowing how to operate in traffic. This is a lie, plain and simple. Nobody has ever demonstrated that cyclists can safely ride in a useful manner without knowing how to operate in traffic. None of the facilities that Wilkinson advocates enable this to be done, and Wilkinson has not even tried to actually demonstrate this claim that he has so frequently made.}

"Increasing the number of users: This policy calls for assessing the facility needs to determine what would be required to encourage more people to use bicycles. This policy establishes a more demanding performance measure by which to determine the success of any actions, namely that the provision result in a real increase in bicycle use. Increasing bicycle use means attracting new users and those new users will often not be willing to share roadway lanes with motor vehicles under existing traffic conditions."

{When considered in the context of Wilkinson's previous paragraph on accommodating present use, the meaning of this statement is that Wilkinson is advocating a dangerous and unethical policy. That is, his prime directive is attracting non-cyclists to transportational cycling by providing what these people want, people who must be ignorant of the subject, regardless of how dangerous this is for them. That is, Wilkinson wants to attract people who, because they are frightened, hate the idea of learning how to operate in traffic, who believe that bikeways remove the need for that knowledge, and who believe that bikeways protect them from motor traffic even when they don't have the knowledge that they hate. Wilkinson's argument would be acceptable only if bikeways did these things. However, none of these results have ever been demonstrated for any practical system of bikeways; indeed, in most systems the demonstration is that cycling for transportation on bikeways requires greater knowledge and skill than cycling for transportation on normal roads.}

C&T refer to Wilkinson's assertion that few non-cyclists have become "`effective cyclists' through a rigorous training and education program" and they argue that instead we must provide what the non-cyclists want.

{This statement is incorrect in two ways. First, taking an Effective Cycling course is only one way of learning vehicular-method cycling. Several hundred thousand copies of Effective Cycling and similar books (more of these than of the original) have been sold, generally to people who have been interested in learning vehicular-method cycling. These people have passed many of their skills on to those with whom they ride. Experience itself produces the change. While we know of many cyclists who have progressed from the state of fear to competent and confident cycling, the only ones who seem to have regressed are a very few militant bikeway advocates who have their own reasons for hating vehicular-method cycling.

Second, Effective Cycling is purely a private program that runs against the full weight of governmental policy and popular superstition. Achieving any substantial results under these conditions shows the enormous superiority of the vehicular-cycling method. Had government put its resources behind what is the only known method of cycling safely for transportation, instead of its contrary bikeway policy, the results would surely have been greatly different.}

C&T mistakenly describe as "current research" Wilkinson's classification of cyclists as child, basic, and advanced. {There's no research in that classification; it is merely Wilkinson's opinion. The purpose of Wilkinson's classification is to separate the cyclists whom he calls "experienced" and whom C&T call "elitists" from other cyclists. They use this classification to deny the value of experience in developing knowledge and skill. They assert that instead of listening to those who know, society must provide what the inexperienced and ignorant believe to be necessary.}

2.4.3 Separation versus Integration

C&T assert that Wilkinson's ABC classification obsoletes the argument about whether cyclists should be "integrat[ed] into the traffic flow and train[ed] to ride in a vehicular manner." In support they argue that the argument has been carried on by two extreme groups, the Elite Cyclists with their vehicular-cycling method and book and the Bicycling is Dangerous Public with its demand for separation.

{Describing the vehicular-cycling method as extreme is an absurd lie. It is the only known safe method for performing cycling transportation anywhere except in a few old-world urban centers where distances are very short. Branding those who recognize its value as Elitists is also a lie. The vehicular-cycling method can be practiced by almost anyone because it requires no more than ordinary physical and mental characteristics. It has been adopted as the standard training for people who have been refused motor-vehicle driving licenses on the basis of mental incompetence. Those who advocate vehicular-method cycling know that it is best for everybody beyond childhood. The idea that vehicular-method cycling requires extremely rare physical or mental abilities is merely another demonstration of the exaggerated fear of traffic.(4) When people believe that cycling in traffic is extremely dangerous, they naturally conclude that the ability to stay alive in that traffic requires superhuman abilities.

While C&T also call the Bicycling is Dangerous believers extremists, they still defer to that belief. Their recommendations are those of the Bicycling is Dangerous group, which are that same-direction motor traffic is the prime danger to cyclists, who must be protected against it by bikeways, the facilities that make it safe to ride without attaining competence. While C&T call that extremism, they still make those recommendations.}

C&T argue against what they call the Elite Cyclists, erroneously saying that the vehicular-cycling response developed from "requirements to use separate bike paths ... [and] diminution of a bicyclist's right to the road."

{The question of whether vehicular-method cycling is elitist is discussed above. C&T are only partially correct in writing that the restriction of cyclists' rights was a reason for developing the science of vehicular-method cycling. The reason for the conflict is that cyclists knew that vehicular-method cycling was the best method known. If bikeway cycling were preferable, there would have been no reason for bringing up the denial of rights, for people don't bother to protect rights that provide no benefit for them. The reason for bringing rights and their legal protections into the discussion at all was to protect the best and proper method of cycling against the governmental and societal efforts to restrict and prohibit it.}

C&T also argue that today's bicycle facilities are much better than those designed according to the original standards. "For many people, therefore, the debate between separation and integration is looking increasingly contrived. Planners and engineers are rejecting the notion that all special bicycle facilities are unnecessary and dangerous, as some believe, or that any type of bike facility is good, as others contend. There is substantial middle ground in the debate."

{Here we have two straw men battling each other. The question always has been which type of facility best serves cyclists. However, we need to recognize that in most test cases a well-designed normal roadway has been shown to be the best facility.}

2.4.4 Designing and Selecting Facilities

C&T assert that "the safety of cyclists is influenced more by the design of a particular facility than the decision to use [they mean install] that type of facility." They list the following types: shoulder, wide curb lane, bicycle route, bike lane, bike path, shared lane, bicycle and bus lane, bicycle boulevard, and traffic calmed neighborhood. "Each of these facilities can accommodate bicyclists without compromising. Each of these facilities can also enhance the safety of bicyclists and/or increase bicycle use." They then have a section for each of these.

{Their claim that none of these types of facility is more dangerous than a well-designed normal roadway will be discussed below.}

2.4.5 Shoulders

While C&T state that shoulders provide benefits to motorists in the absence of cyclists, they never mention the benefits of shoulders in making it easy to overtake cyclists. While they say that shoulders benefit cyclists, they never describe how they do this.

2.4.6 Wide Curb Lanes

C&T claim that "wide curb lanes accommodate only the group A cyclist. ... However, they are successful in increasing the capacity of highways, especially at intersections." {What C&T mean is that B cyclists don't understand that wide curb lanes work perfectly well for them. The problem is not the physical design, but the ignorance and superstition that a stripe is also necessary, in part promoted by government.}

2.4.7 Bicycle Routes

C&T quote an article in Pro Bike News for the following information: Bicycle routes often don't go where cyclists need to go. ... They don't go where group A cyclists go and they don't provide the "sense of protection" that B/C cyclists desire. ... Poor implementation. ... Motorists think cyclists are restricted to bike route streets.

C&T concede that bike routes may be useful as highway and destination markers, if properly located.

{Bike routes don't go where cyclists want to do because they have been selected according to where other people think cyclists should ride instead of where cyclists want to ride. The supposed lack of protection is merely the lack of something imaginary, because the other facilities don't protect cyclists.}

2.4.8 Bicycle Lanes

C&T praise bike lanes, in the midblock situation, for causing motorists to swerve less. They refer to the argument that bike lane stripes reduce the sweeping action of motor traffic. As the prime argument against bike lanes, C&T advance a straw man in the shape of the work of the Lotts, two ill-informed and incompetent bikeway advocates from the early days in Davis, California. The Lotts give the two arguments as follows: "Bike lanes prevent bicyclists from making correct left turns (from the left lane) and motorists from making correct right turns (from the right edge of the roadway)." Then C&T think that they have countered these arguments by quoting the article by Ronkin from Pro Bike News: "The difficulties associated with making a left turn in traffic has nothing to do with the presence of a bike lane. A bicyclist has to be prudent and look over their shoulder repeatedly before merging to the left regardless of conditions. ... Conflicts with right-turning movements are dealt with by dashing the bike lane stripe before intersections and dropping the markings altogether across intersections. ... Do bicyclists really want to be dependent on automobile traffic to blow the street clean?" {Bike lanes do not prevent cyclists from making correct left turns; they merely discourage cyclists from doing so and encourage motorists to believe that they should not do so. The competent cyclist can make left turns regardless of the bike-lane stripe. The problem is that the bike-lane stripe tells the uninformed cyclist, just the type of cyclist that is supposed to benefit from the stripe, that he should not cross it, or should cross it as infrequently as possible and as late as possible. Observation shows that in bike-lane cities, cyclists make many more left turns from the side of the road without looking behind. As far as motorist right turns are concerned, motorists can merge across the bike-lane stripe to make right turns, but the evidence is that they frequently do not do so because they believe that the bike lane is for cyclists only.(5) As for street cleaning, cyclists fare better riding on surfaces that are cleaned automatically by motor traffic than on surfaces that might get cleaned once a month.}

C&T quote Dan Burden as giving these additional advantages of bike lane stripes: the "allow motorists to turn into the roadway without encroaching on another lane, and negate the need for wide-turning radii at corners, thus maintaining shorter crossing distances for pedestrians at intersections." {This is nonsense. Bike-lane stripes do none of these things. As they so frequently do, C&T are confusing the effect of wide roads with the effect of bike-lane stripes.}

C&T state that "opponents of bike lanes ... question the safety of bike lanes." They first assert that this argument came from followers of Forester rather than Forester himself. {This is false. I discussed this in Effective Cycling and in greater detail in Bicycle Transportation, and I supported my analysis with data from the observation of many cyclists. I observed several thousand maneuvers by cyclists, of which 1,365 were of types that showed statistically significant differences of behavior between cities with bike-lane systems and cities without bike-lane systems. The observations showed much higher proportions of dangerous maneuvers by cyclists (and by motorists to, although not counted) in cities with bike-lane systems than in cities without.(5) C&T don't know the literature on their subject.}

To counter this argument, C&T quote several studies. The first is that of Kaplan, who they say "actually found bike lanes to be slightly over half as dangerous as major and minor highways." {This is false. Kaplan's data concerned the year 1974 and his classification was titled "On-Street Bicycle Facility (Lanes, Routes)." Considering the very small number of bike lanes in 1974, most of this cycling was on normal roads, largely in rural areas, specially selected for their safety.(6)} They then quote the analysis of the Santa Barbara traffic engineer who, in 1976, claimed [C&T use the word "reported"] that bike lanes would have prevented 14% of car-bike collisions. {This is merely the result of a preliminary analysis of car-bike collisions, by an engineer who had had no experience with bike lanes, on the basis of claims for bike lanes for which, at that time, there was no evidence whatever.} C&T claim, on the basis of a city report, that installation of bike lanes on 18th Ave. in Eugene, OR, in 1979, reduced the "bicycle accident rate" by from 40% to 70%, and reduced the motor-vehicle collision rate also. {This was not the result of merely installing bike lanes. Many more things were done, which would reasonably account for nearly all of the reduction. As elsewhere, C&T use `bicycle accidents' or `bicycle crashes' to describe only car-bike collisions, ignoring that the great majority, >80%, of accidents to cyclists are not car-bike collisions.} They quote Ronkin again about a newspaper story, dated 1982, that bike lanes in Corvallis, OR, reduced "bicycle crashes" by 60%. {The article described other actions, including widening roads, removing parking, and community activism that more likely account for the reduction in accidents. The article, and a letter from the Corvallis Public Works Department supporting it, are so deficient in data that no strong conclusions can be drawn, but in the year after the accident the accident rate per street-mile on the streets with bike lanes was about 6 times higher than on the streets without bike lanes. That ought to raise some question about the accuracy of the study and the credibility of C&T's conclusions.} The only supposedly scientific study that C&T quote is one by Smith and Walsh from Madison, WI, which concludes merely that, although the accidents on the bike-laned streets increased somewhat, "the bicycle lane corridor accidents did not increase significantly."(8) C&T call "more dramatic" the conclusion of a Danish study published by the Bicycle Federation of America (a bikeway advocacy group that employs C&T) which demonstrates merely that bike lanes are no more dangerous than bike paths.

{The evidence that C&T develop for the safety of bike lanes consists of one study that did not cover bike lanes, one prediction before anything was known, one report to a city council about a project that covered more than bike lane stripes, one newspaper story with the same characteristics, and one published study. While the data of that study showed an increase in bicycle accidents, the authors concluded that the increase was not statistically significant. This is very little and sketchy evidence considering the extent to which bike lanes have been installed since 1976 and the claims that are made for them. The Danish conclusion that bike lanes are no more dangerous than bike paths, the most dangerous facilities known, condemns bike lanes rather than exonerates them. C&T continually ignore the question of whether the supposed claims for bike lane stripes come from the stripes themselves or from other actions commonly done at the same time, such as other restriping of the street, widening of the street, removal of on-street parking, and the like.}

Finally, C&T argue that non-cyclists want bike lanes because they believe that bike lanes protect them from cars. {As you can see from the above discussion, there is no basis for this superstition.}

2.4.9 Bicycle Paths

C&T first argue against the view that bicycle paths are dangerous. The only study of the accident rate of cycling on bike paths is that of Kaplan, whose data show that cycling on bike paths has an accident rate 2.6 times that of the average for roads. C&T denigrate the Kaplan study as being old, based only on experienced cyclists with few accidents. {The Kaplan study covered 3270 cyclists who rode over 7 million miles, and incurred 854 accidents.}

C&T quote Wilkinson's argument that "based on this single source, Forester articulated the theory that separate bicycle facilities are less safe than vehicular-cycling on the roadway." {This is false. I had long before made the engineering analysis of how separate bicycle facilities adversely affect the safe and proper operation of bicycles. The data from the Kaplan and the Cross studies merely provided statistical authentication for the theory that had been thoroughly worked out long before. Subsequent to these studies, the steady stream of complaints from experienced cyclists about the dangers of bike paths, and the frequent enactment of very low speed limits for popular paths because of the dangerous nature of both their intersections with roads and of the traffic on them, give further confirmation of the accuracy of the original studies. Scientists consider it strong confirmation when independent data confirm a previously stated hypothesis.}

C&T admit that bike paths parallel to city roadways are dangerous and that these facilities gave bike paths a bad name. They then describe the phenomenal growth of the Rails to Trails Conservancy, but do not recognize that because of the small number and restricted locations of the possible routes such routes can never supply more than a minute portion of the need for cycling transportation. C&T recognize that any such facility is going to be a multi-use path, but they do not list the disadvantages of such operation. Finally, they write that "trails ... are good places to learn the skills and techniques that are necessary when riding in traffic." {This latter claim shows C&T's total ignorance of the skills and techniques that are necessary for cycling in traffic. If this were not tragic it would be laughable.}

2.4.10 Shared Lanes

C&T state that on low-volume roads "all adult and child bicyclists can operate safely and without being intimidated by traffic." C&T describe several systems for evaluating the suitability of roads for cycling. They all end up by considering only the factors about the intensity of same-direction motor traffic. One explicitly says that roads with much same-direction motor traffic are dangerous. {I have recently reviewed some of these studies. They have ended up by being merely subjective judgements by ill-informed people of the intensity of the fear of same-direction motor traffic that they feel on each street.(9)}

2.4.11 Bicycle and Bus Lanes

C&T describe these as wide outside lanes that are limited to bicycle and bus traffic. C&T do not describe the advantages these provide for cyclists, saying only that cyclists like these. These would only be installed where there is significant bus traffic. Providing space so that cyclists can lawfully ride to the left of buses is an advantage. However, if buses swing out to overtake other buses stopped at bus stops, behavior which one would expect in locations where there is much bus traffic, then the advantage may well be lost and cyclists would fare better if they cycled further left.

2.4.12 Bicycle Boulevards

C&T describe these, but the reader has to guess from their description that this provides the speed of cycling on arterial roads with the low traffic of residential roads.

2.4.13 Conclusions

"The most important development in the last decade is that the separation vs integration debate has become redundant and futile. Enough is known about the design and appropriate use of different bicycle facilities and shared roadways to counter any fears that one type of facility is inherently more or less dangerous than another. The 1992 FHWA report [The Effects of Bicycle Accommodations on Bicycle/Motor Vehicle Safety and Traffic Operations] is a major breakthrough in this regard, highlighting the realization that different facility types can achieve markedly different outcomes in terms of level of bicycle use ... all within acceptable margins of user safety." {I discuss these conclusions in my section titled analysis and conclusions.}

2.5 Surface Quality

C&T write that not much is known about the proportion of accidents to cyclists that are caused by defective surface conditions. C&T mention railroad crossings, drainage grates, gutter-edge ridges, and cobblestones [not by that name]. Not much useful information.

2.6 Traffic Calming

C&T describe this as a technique for slowing motor traffic for the benefit of the neighborhood. C&T do not discuss the effect on cyclists. They quote a study saying that average motor vehicle speeds have been significantly reduced although average trip times have increased only by 33 seconds. {For reasonable speed reductions, the maximum trip distance for which the increased time is only 33 seconds is less than 2 miles. This is not very pertinent to American conditions.}

2.7 Safety Equipment

2.7.1 Lights and Reflectors

C&T follow the usual superstitions in discussing this subject, and shows little sign of understanding it.

2.7.2 Helmets

C&T favor programs to encourage wearing of helmets, and mentions mandatory helmet laws, but show no awareness of the debate about the effectiveness of helmets in preventing brain injury and of helmet laws in reducing cycling.

2.8 Education

C&T discuss several types of program: short articles and videos; stage presentations; in-school teaching with bicycles but not on road; bicycle rodeos; Effective Cycling training, primarily for adults. C&T express no judgement about these programs, but they write: "What appears to be missing ... are direct observations of bicyclist behavior in the field." {C&T appear not to know that every Effective Cycling course has a final examination that includes direct observation of the behavior of the student cyclists, and that the pilot tests of EC for children included measurement of the behavior of many students. Similarly, C&T appear not to know that EC has been tested with children of ages 8, 10, and 13 years. (10) (11)}

2.9 Enforcement and Regulations

C&T describe the high rate of violations of traffic laws by cyclists and the low priority given to enforcement against them. They also describe the low level of enforcement against motorists who have been involved in car-bike collisions. They believe that the rate of violation by cyclists today has increased from some lower level in the past. They suggest enforcement against the most dangerous violations, appropriate punishments, training of police, and notes that programs of police on bikes are beneficial for cyclists as well as good for other law enforcement purposes.

2.10 Conclusions

We know little about the actual use patterns of cycling transportation and cycling recreation. Crash data is better known. The standards for bicycle facilities are adequate. Deciding which type of facility best suits each circumstance is now well understood. Traffic calming may have great potential. Traffic law enforcement needs consistent implementation. We need better information about the effects of helmet laws.

3 Analysis and Conclusions

3.1 Description of Document

As stated in the introduction this is not a synthesis of safety-related research. Instead, it is intended to persuade a reader to believe in the rightness of the cycling program and policy that are desired by the Federal Highway Administration and its hired creature, the Bicycle Federation of America. The document probably persuades the intended readers, who are politicians, highway administrators, and bike planners. However, it is so biased and so demonstrative of ignorance that it persuades a well-informed reader to be very suspicious of any program or policy that is advocated by such bias and ignorance, and of the organizations that produce such documents.

3.2 Authors' Intents

C&T had several linked intentions, all of which are associated with the great cycling debate. That debate concerns the best method of cycling in urban areas. Should cyclists act like drivers of vehicles, or should they act in some inferior way under the supposed protection of bikeways?

I doubt that C&T had a precise concept of their intentions, but rather I consider that they were following the systematic superstition that controls their thoughts. This superstition is that cycling in traffic is very dangerous and requires either extreme cycling skill or the protection of bikeways. The alternative explanation is that the authors know the truth but chose, for their own purposes, to write according to the superstition. I prefer the explanation of ignorance and incompetence governed by superstition rather than the explanation of deliberate systematic deceit for the reason that if the authors really knew the subject and intended to mislead the reader, they would have made a more persuasive job of it.

  1. One intent was to oppose the vehicular-cycling method, which they did only by denigrating it as elitist.

  2. Another intent was to demonstrate that all types of special bicycle facility can be just as safe as normal roads.

  3. A third intent, also linked with the second, was to continue having the reader believe the superstition that these special bicycle facilities protect the cyclist from motor traffic.

  4. A fourth intent, linked with the second, was to convince the reader that in a city with special bicycle facilities there is no need to learn the proper method of cycling, that one can safely cycle to any destination without that knowledge.

3.3 Opposing the Vehicular-cycling Method and Effective Cycling

C&T had to oppose the vehicular-cycling method and those who advocate it because those are the only principled opponents to their program (conservative attitudes, laziness, and tax-saving, while effective, are not based on principles of cycling). The only direct objection that the authors even tried to make is their statement that this method is elitist and has attracted only 3,000 users.

Both claims are false. Vehicular-cycling technique is practiced by many more than the number who have taken Effective Cycling classes. The Effective Cycling Program is merely the formalization of the technique that has been developed and practiced by cyclists for decades. Many experienced cyclists who read Effective Cycling for the first time say that they had been doing most of it for years, but they had never before read or heard a formal explanation and justification for their actions.

Vehicular cycling technique is not difficult to learn and can be learned by quite young children. Classes of children of 10 years of age can in fifteen hours learn the skills of following all of the five principles of cycling in traffic, and when they do so they are much better than the average adult cyclists around them. Something that is so easy to learn cannot be an elite skill.

3.4 Claiming that all types of facility are safe

C&T claim that all types of bicycle facility, when designed in accordance with the standards and when properly located, are safe. They do this to preserve the authority of government and bike planners to design the bike planned city. If some facilities were dangerous, then those would have to be disallowed. This really applies only to bike paths and, to some extent, to bike lanes. Shoulders, wide curb lanes, bus and bike lanes, bicycle routes, and bicycle boulevards have never been subject to the charge that they endanger cyclists. Bike paths are of two types: that alongside an existing road and that far distant from roads. C&T accept that bike paths alongside city streets are dangerous. While C&T denigrate the data showing that bike paths are dangerous, they list the traffic that makes them dangerous, and they advance no reasons for believing that bike paths well away from roads are safe. As for bike lanes, the reports of accident reduction that they quote are old, not done as scientific investigations, and fail to separate the effects of the bike-lane stripe from the other improvements made. Of the only three scientific studies that they quote, they misquote one to say that its data was derived only from bike lanes, another concluded only that bike lanes did not significantly increase accidents, while the last damned bike lanes with the faint praise of being no more dangerous than bike paths alongside streets. If these are the only pieces of evidence supporting the safety of bikeways that can be advanced by bikeway promoters assisted by governmental research, then the case for the concept that bike lanes and bike paths are as safe as normal streets is hardly persuasive, and the opposing data and analysis that holds the opposite should be believed.

3.5 Claiming that bikeways protect cyclists

Aside from their claims about the two bike-lane projects that, so they say, reduced accidents to cyclists, C&T never explicitly claim that bikeways protect cyclists from motor traffic. However, they explicitly state that the public believes that bikeways protect cyclists from motor traffic, and they never question that belief. So far as the general public is concerned, the entire function of bikeways is to protect cyclists from the dangers of same-direction traffic. C&T write that "the incidence of these crashes is relatively low," but they still overstate the proportion by including car-bike collisions caused by other factors, such as cyclist swerves and defective nighttime equipment. In urban areas in daylight only 0.2%-0.3% of accidents to cyclists are caused by motorists overtaking a lawful cyclist.(12) In short, bikeways cannot protect cyclists to any significant extent because the danger against which they are supposed to protect only infrequently causes accidents.

3.6 Ignoring Intersections and Other Places Where Traffic Streams Cross

A very large portion, approximately 90%, of car-bike collisions occur at intersections or other places where traffic streams cross. Any attempt to reduce car-bike collisions must consider the appropriate designs of such places, and the appropriate design must be based on the appropriate driving method. C&T's discussion of appropriate intersection design is cursory and inaccurate, often recommending dangerous designs. Without knowledge of proper driving methods, they cannot provide appropriate information. This set of errors demonstrates the effect of the superstition that same-direction traffic is the prime danger to cyclists (when cross traffic is a far greater danger) and of the political desire for a bikeway-based cycling program.

3.7 Claiming that bikeways eliminate the need for cycling skill

C&T never explicitly claim that bikeways eliminate the need for cycling skill, but they sneak that concept in repeatedly by saying that bike lanes make streets suitable for beginning cyclists and that bike paths are wonderful places for learning traffic cycling. However, they present no analysis of cycling skill. Naturally, without that they cannot demonstrate which parts of that skill can be eliminated in cities with bikeway programs. The claim that bikeways eliminate the need for any part of cycling skill has no basis whatever.

Naturally, C&T make no mention of the analyses that have been made of cycling skill. The analysis that merely described the normal traffic-cycling skills demonstrated that all of these would be required in any case, while that which compared the skills required for normal streets with those required for bikeways demonstrated that more skills were required for cycling in cities with bikeways.(4) (13) (14)

3.8 Items Absent from Document

Although this document calls itself a synthesis of the results of safety-related research on cycling, its authors admit that it includes much other material that suits their agenda. At the same time, this document completely ignores many important safety-related subjects whose results have been developed through research. Among the items absent are:

  1. Methods of operation

  2. Analysis of accidents

  3. Distinguishing between direct and confounding causes and effects

  4. Development of skills

  5. Differences in urban design

3.8.1 Methods of operation

More than merely ignoring the methods of operation, the methods of cycling in traffic, C&T summarily dismiss this knowledge by calling it elitist. It is impossible to rationally discuss the effects of different designs of facility without discussing how they are to be used. Discussion earlier in this paper has discussed the fact that bikeways do not reduce the need to know the methods of traffic cycling and that bike lanes and urban bike paths require, for safe operation, a better understanding of the principles of operation because they present more difficult situations.

We have detailed knowledge of the driving method used by typical cyclists. We have detailed knowledge of the driving method used by competent cyclists. We have even more detailed knowledge of this driving method as it is used by drivers of motor vehicles. We have detailed analyses of the relationship of all of these driving methods to the design of the facilities and the design of the traffic laws which control the method of use. C&T have ignored all of this knowledge.

It is reasonable to conclude that the authors failed to include any discussion of driving methods because any such discussion would demonstrate the falsity of their agenda. However, I think that this was not so much a deliberate omission as it is the result of bicycle advocates' long-standing practice of ignoring the subject by presuming that it concerns only elite knowledge. That is, the opposition to bikeways of those who recognize the value of the vehicular-cycling method has resulted in what is now the bikeway advocates' automatic response that this must be elitism.

3.8.2 Analysis of accidents

C&T make the claim that two installations of bike lanes dramatically reduced accidents to cyclists, by values of 60% and 70%. Anyone with knowledge about accidents to cyclists should know that less than 1/6 of accidents to cyclists are car-bike collisions. Anyone with knowledge of car-bike collisions should know that only a small fraction are of the types that urban bikeways would prevent. Therefore, the authors should have recognized that these two claims are preposterous, even if they intend the reduction to consider only car-bike collisions. Bike-lane stripes could not have produced this difference.

C&T have completely ignored discussion of accident types and causes. Without knowledge of the different types of accident and their frequencies (even if only relative frequencies), and knowledge of their relationship to the driving methods, there is no possibility of rationally discussing the safety effect of any facility design. Any reasonable discussion of accidents to cyclists discloses that even if bike lanes or urban bike paths operated perfectly, they could prevent only a small proportion of car-bike collisions, and that they probably produce a large number accidents as adverse side effects.

There are two reasonable explanations for the authors' refusal to discuss this subject. The first is that they unconsciously assumed that the prime danger to cyclists is same-direction motor traffic, an assumption that seemed so obvious that there was no need to discuss it. The other reasonable explanation is that the authors were aware of the importance of this subject and chose to ignore it on the estimate that those whom they wished to influence by their report would make that assumption without questioning it. The authors did mention the subject once, showing that they have some inkling of its importance.

Although C&T made the two preposterous claims about the amount of accident reduction produced by bike lanes, they seem to have largely given up the claims that bikeways markedly reduce accidents. They have reduced their formal claim to the one that bikeways, when installed in the proper locations, are no more dangerous than normal roads. With the safety issue thus pushed aside, they think that it becomes legitimate to rate bikeways by the number of non-cyclists that they immediately attract. The problem is that the safety issue, which has long-term consequences as well as short-term ones, has been eliminated only by the authors' superstition or mendacity, not by scientific knowledge, and it remains a very pertinent subject for discussion and analysis.

3.8.3 Distinguishing between direct and confounding causes and effects

There has never been a scientifically reasonable test of all of the effects of bike lanes. C&T ignore this problem; they probably do not understand that it exists. In summary, nobody has performed tests that separated out the effect of the bike-lane stripe itself. There have been many cursory studies, often politically motivated, of the results of doing a great many things in addition to painting bike-lane stripes, with the results reported as the result of the installation of bike lanes. Typically, a city starts with a controversial political discussion of bicycle safety, as a result of which it has a bike-safety campaign, part of which takes a narrow road with on-street parking and converts it to a wide road without on-street parking and with a bike lane stripe, and reports the resulting change as the result of painting the bike-lane stripe.

A street must have either a wide outside lane or a well-paved shoulder before a bike-lane stripe can be painted. Analysis of accidents and driving methods shows that practically all of the possible improvement (both in safety and in capacity) is produced by the wide outside lane or the well-paved shoulder and that the painted stripe offers little, if any, further improvement. However, the whole job gets done at once and the improvement is reported as produced by the painted stripe.

In a public debate in San Francisco among some of the best-informed cyclists in the nation, both bikeway advocates and bikeway opponents, it was agreed that the bike-lane stripe itself has no real safety effect and is merely a political device. That is, because the public believes the false superstition that the bike-lane stripe "makes cycling safe," the public appeal to that superstition can develop the political power to do the things that actually produce improvement.

C&T either are ignorant of a subject that has been extensively discussed in this field or they are sufficiently mendacious to conceal their knowledge because discussing the subject would jeopardize their agenda.

3.8.4 Development of skills

C&T's only discussion of traffic-cycling skill summarily dismisses it as elitism that is practiced solely by a hard core of present transportational cyclists, largely influenced by Effective Cycling. That dismissal leaves out much more than the present level of skill. The authors tacitly assume that if we develop a reasonable level of cycling transportation the future participants will remain as ignorant of cycling as the present-day non-cyclists whom they wish to attract. On the face of it, this is absurd; by practicing any activity one's skills improve.

The evidence that exists, from both the USA and Great Britain, shows that cycling experience reduces the accident rate by about 80%.(15) (16) In other words, even in cycling, skills naturally develop enormously with experience. There is plenty of anecdotal evidence from readers of Effective Cycling who have commented that they had already worked out, on their own, from their own experience, much of the driving method described therein. What surprised them was not the methods described, but that this was the first time that they had read descriptions of these methods and, more frequently, the first time that they had read the scientific justification for them.

Effective Cycling contains few new methods. Almost the only new methods concern multiple turn lanes and traffic circles, designs that are relatively new. The rest is old material, merely the formalization of methods that had been practiced and tested over decades. All that Effective Cycling does is to enable the user to learn rapidly what he or she would tend to learn naturally over a much longer period.

The person who practices transportational cycling will, unless the learning is interfered with (more about this later), gradually learn the skills and methods that are described in Effective Cycling, in which case he or she fits C&T's definition of a Group A elite cyclist. It would be to the advantage of society if that person learned early, as by training, than if that person learned over a longer time by trial and error, incurring the dissatisfaction and injuries produced by that process. There is plenty of evidence, both analytical (based on principles of human performance) and statistical (based on the Effective Cyclist driving test results), that quite ordinary people have no difficulty in learning the proper methods. As well, the methods are very similar to driving a motor vehicle and use the same human abilities, and in the USA we figure that practically everyone can learn to do that in reasonable time.

As before, C&T have ignored this well-discussed and important subject. As before, they may have ignored it because they are ignorant of it, or they may have mendaciously omitted it because it would jeopardize the policy and program that their organizations prefer.

3.8.5 Differences in urban design

C&T quote several other authors to the effect that we will not have significant cycling transportation in the USA until we redesign our cities to reduce the travel distance. When making transportational trips, many people choose the faster mode unless the difference in time is insignificant. One very powerful reason why so many Dutch people choose to cycle for so many of their trips is that motoring conditions are so difficult that cycling is both easier and quicker, door to door, than motoring. Shorter travel distances reduce, even cancel, the time advantage of motoring over cycling. While C&T do not discuss methods of redesigning our cities, they support the concept.

However, this concept raises two considerations. Increasing the speed of cyclists reduces the travel time in the same way as does reducing the travel distance. American transportational cyclists ride much further and much faster than do typical transportational cyclists in places like Amsterdam. Riding fast requires that the cyclist operate in the vehicular manner on well-designed roads. Typical urban bike paths are very dangerous when used in this manner. Bike lanes may not be so speed sensitive, but they have other disadvantages. In general, the cyclist who refuses to operate in the vehicular manner must operate either slowly or dangerously. C&T omitted all discussion of the speed of cycling, although this has been well reported. Had they discussed this, the discussion would have jeopardized the policy and program preferred by their organizations.

The second consideration is that of the length of time it will take to redesign our cities to significantly shorten travel distances. In the USA our cities have grown since 1920 to accommodate a large increase in population, and they have largely grown around the transportation opportunities offered by the private automobile. (Read The Great Gatsby or Only Yesterday(17) for pictures of how rapidly the private automobile took over urban growth.) That was seventy years ago. With a presumably much lower growth rate, how long would it take to rebuild our cities so that, say, half of the congestion-making trips become less than two miles? That will take decades, if it occurs at all. To try to do it faster would mean abandoning much of our building stock, residential, commercial, and possibly even industrial. In the meantime, cyclists still have to ride the present distances, which requires vehicular-method cycling.

At the present time, there is a process of reducing travel time by producing satellite urban centers, "Edge Cities," to use the title of a book.(18) These have become economically viable because of the difficulty and cost of getting to and operating in downtown. However, these centers are dispersed at motoring distances, not walking distances, so that cycling between them still requires fast cycling to be a desirable choice.

3.9 Two Different Policies

My discussion of C&T's document has introduced the question of different policies and programs for cycling transportation, stating that C&T chose to support the one preferred by those who pay them, the Federal Highway Administration and the nation-wide bicycle programming system. There is another policy and program that makes directly opposite recommendations and ought to achieve very different results.

3.9.1 The Bikeway Policy

The governmental policy and program is one of bikeways that used to be based largely, so it was said, on bike-safety claims. In fact, it has always had a large component of motorist convenience, getting cyclists out of motorists' way. Since there has been no scientific study demonstrating that bikeways "make cycling safe," to use the colloquial term (and very little likelihood that such an unlikely proposition could ever be demonstrated), bikeway advocates have turned to the popularity argument. They argue that by giving non-cyclists what they desire, they will produce a system of cycling transportation that will significantly reduce the use of the private automobile, with many benefits to society.

The policy that they advocate is to use the popular fear of cycling in traffic and the popular superstition that bikeways protect against motor traffic and "make cycling safe." Their program is to develop a system of bikeways which will attract non-cyclists to the kind of cycling transportation that non-cyclists understand. That is, short-distance cycling without learning how to operate in traffic.

The are difficulties with this policy and program. There will be greater public demand than exists today for bike paths, the type of facility that the public believes most strongly protects them from motor traffic. This is also the most dangerous type of facility, but the policy cuts off all possibility of telling the public the truth. This is also the type of facility that is least useful because it is least convenient and requires the slowest speeds. While safe operation in bicycle lanes requires that users (both cyclists and motorists) operate in the vehicular manner, the policy says that it is safe to operate in bike lanes without learning either general safe operation or the special additional skills required to avoid the difficulties with bike lanes. If the requirement to learn safe driving practice is admitted, then the justification for the lanes practically disappears.

In summary, the more that facilities are designed around the public superstition, the greater and stronger are the social forces that support that superstition, both the strength of its organizational existence and the strength of its effect on the public. While this may initially produce more cyclists, their cycling will be done under conditions of less usefulness and greater danger. Such a policy and program are typical of Dutch practice, and Dutch cyclists don't rebel at their conditions. However, they also do not use cycling transportation when their local conditions are like typical American ones. It is reasonable to conclude that a program that results in slower cycling, for shorter distances, with less convenience, and with greater fear and danger, when tried in the USA, in the long run will produce little cycling transportation.

Some argue that this is a useful starting policy and program, from which we will naturally develop into the program that is described below. I see no evidence for this view; the Dutch have had their system for about forty years and show little signs of change. When a system is as firmly established as theirs, it requires practical rebellion by the users to change it. Although there are some signs of discontent among the better-informed Dutch cyclists, there is nothing like the rebellion that would be required to change the system.

The bikeway policy and program are based on the public's fear of same-direction traffic, which although inaccurate is very strong. The establishment of a system of facilities and procedures that depend on that fear strengthens that fear and makes any contrary thought seem that much more outrageous, or as the bikeway advocates put it, "elitist." The more that is done in this direction, the harder it will be to change later.

3.9.2 The Cycling Transportation Policy

The contrary policy is to tell the public the truth about cycling transportation as it affects the user. That is, that "cyclists fare best when they act and are treated as drivers of vehicles." Also, we must teach the public the falsity of the contrary superstition, that cycling in the standard manner is dangerous and beyond the abilities of ordinary people. On the contrary, all the evidence shows that all road users must follow the same driving methods, which we know are humanly possible and reasonably safe.

It is impossible to operate safely and efficiently, in any system, without having the skill of riding in traffic. That skill is reasonably easy to learn, and once cyclists have learned that skill their cycling is reasonably safe and very useful.

While there are many improvements that accommodate such cycling, bike paths and bike lanes are not among them, for the reasons given above. The improvements that work are those that do not upset normal operation on the roadway by catering to the belief that a different way of driving is safer. Bike lanes fail this test, as do bike paths where they contact roads. There are many improvements that make cycling easier: shoulders or wide curb lanes (these are the same as far as cyclists are concerned), traffic signals that respond to bicycles, right-turn-only lanes to separate straight-through cyclists from right-turning traffic, protected left turn signal phases and lanes, drain grates that don't catch wheels, more space at diagonal railroad crossings so cyclists can turn perpendicular to the tracks, better bicycle parking facilities, and more.

Such a program is likely to be considerably slower in initially attracting cyclists, but it will operate with a lower accident rate, a higher level of satisfaction, greater usefulness, and produce more cycling transportation per user. In the long run, such a system is far better than one that is limited by its own superstitions to operation that is slow speed, short distance, inconvenient, dangerous, and frightened.


1 Forester, John; Effective Cycling, 6th edition, 1983, The MIT Press

2 Forester, John; Bicycle Transportation, 2nd ed., 1984, The MIT Press

3 Bicycle Transportation, Chapter 13

4 Effective Cycling Chapters 26-32

5 Forester, John; The Effect of Bike-lane System Design on Cyclist Errors; 1978, 1982; Custom Cycle Fitments, Sunnyvale CA

6 Kaplan, Jerrold A.; Characteristics of the Regular Adult Bicycle Rider;1975; Masters Thesis, University of Maryland.

7 Cross, Kenneth D., & Gary Fisher; A Study of Bicycle/Motor-Vehicle Accidents: Identification of Problem Types and Countermeasure Approaches; 1978; National Highway Transportation Safety Administration; Washington DC

8 Smith & Walsh; Safety Impacts of Bicycle Lanes; Transportation Research Record 1168; Transportation Research Board, Washington DC.

9 Forester, John; Studies on the Opinions of Cyclists about Different Street Conditions; 1996

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

11 Forester, John; Elementary-Level Cyclist Training Program: Objectives, Techniques and Results; 1982

12 Calculated from Cross's and Kaplan's statistics. See Bicycle Transportation Chapter 5

13 Bicycle Transportation Chapter 9

14 Bicycle Rider's Task Analysis; unpublished paper, based on the methods used for the Motorcycle Rider's Task Analysis.

15 Comparison of Kaplan's statistics with those from the National Safety Council's two studies of Bicycle Accidents and Usage Among Elementary-School Children and Among Young Adults; Authors are Chlapeka, Schupack, Planek, Klecka, & Driessen

16 Watkins, S. M.; Cycling Accidents; 1984; Cyclists' Touring Club; Godalming, UK

17 Allen, Frederick Lewis; Only Yesterday: An Informal History of the Nineteen-Twenties; 1931; New York

18 Garreau, Joel; Edge City: Life on the new Frontier; 1991; New York


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