For decades competent and trained cyclists have had to deal with the occasional ignorant motorist who hollers out some variation on “Get off the damn road ya idjyit, yer gonna get yerself killed!” Fortunately most of these incidents don’t escalate into anything serious, and while little has been done to change this attitude over the years, we could shrug off such blather with confidence that the logic of our road use was supported both by direct personal experience and objective scientific study. We also had the backing of the League of American Bicyclists to defend our rights, if only in words.
Well, things have changed. With its recent sophomoric “study” of fatal bicyclist crashes and subsequent statements by its executive director, the League has joined the chorus of ignorance in agreeing that bicycling on regular roadway lanes is “gonna get you killed.”
Fear of the overtaking motorist has been a key arguing point between those supporting separated facilities as the primary accommodation for bicyclists, and those who prefer bicyclists be treated as normal vehicle drivers.
Last month the League released a report on bicyclist fatalities entitled Every Bicyclist Counts. A key finding noted by the League and other bicycle organizations is that 40% of bicyclist fatalities involved an overtaking motorist. This finding varies significantly from other studies and has caused quite a stir amidst bicycle advocates. The fear of the overtaking motorist has been a key arguing point between those supporting separated facilities as the primary accommodation for bicyclists, and those who prefer bicyclists be treated as normal vehicle drivers.
Most notable in the various reporting following the publication of this report is an interview of League President Andy Clarke by BRAIN (Bicycle Retailer And Industry News) in which he attempts to make a connection between the 40% claim and the increasing desire to provide “protected bike lanes”:
Andy Clark: For the longest time it’s been an article of faith that we should be taking the lane, and that separated bike facilities are unnecessary … well, I think we are grown up enough now to say that’s not the case. Most people feel more comfortable actually having a paved shoulder or a cycle track or having a buffered or protected bike lane, and those things will reduce the fear and the incidence of being hit from behind. And we shouldn’t feel bad or awkward about saying that.
BRAIN: So this data reinforces the movement toward separated facilities?
AC: Yes, it reinforces and validates that approach. There is still a small, tiny percentage of people who think we should not be putting in that kind of infrastructure. They are hanging on to the idea that being hit from behind is not that big of a problem. Well, the data tends to suggest otherwise.
There are still going to be places where there isn’t that kind of infrastructure and you are going to need to know how to cope with that. So we are not diminishing the need for the kind of education programs that we have presented and are going to continue to present. But we are saying that infrastructure can and should play a role in eliminating some of the most serious kinds of crashes.
By undercutting the essential practice of lane control, Clarke is dismissing the key defensive driving strategy taught in the League’s own education program, not to mention that of every serious adult cyclist education program.
With such a statement from the League, this 40% overtaking claim is highly significant for those of us who believe bicyclists are drivers of vehicles and wish to protect our rights to use regular roadway lanes. Lane control mitigates many motorist-caused crashes due to turning and crossing conflicts, as well as overtakings, the majority of which are sideswipes in which the motorist is well aware of the cyclist’s presence. By undercutting the essential practice of lane control, Clarke is dismissing the key defensive driving strategy taught in the League’s own education program, not to mention that of every serious adult cyclist education program. A thorough review of the League’s methodology and analysis is called for.
This review will address a number of important shortcomings of the League’s report, including:
Their method of selecting crashes for assessment
Questions of how crashes are categorized
Key differences between urban and rural crashes
Alcohol & drugs and cyclist use of lights and reflectors — two key factors in overtaking crashes
Bias in Crash Selection
Right up front the League’s methodology for selecting and analyzing crashes has a built-in bias. The report states clearly that:
The majority of the information captured by Every Bicyclist Counts came from newspaper reports (56% of all reported sources), TV reports (25%) and blogs (19%).
Using news reports creates bias at two levels. First off, the types of stories that are most heavily reported are those involving the “good cyclist.” The person who is “just like us,” riding a well-maintained, good quality bike, wearing a helmet and obeying the rules. As one who has been following both media attention and official crash data for 20 years, I can tell you the story of the homeless alcoholic hit while crossing a dark arterial at night without lights or reflectors gets scant attention from the media compared to those “good cyclists.” As the League notes, their study misses about 24% of the crashes. It’s quite reasonable to expect that the characteristics of those missed fatalities vary significantly from those they reviewed.
Next, the fact that the League’s sources are from the media rather than the official crash report means the facts have been filtered through a reporter, or a number of reporters. While we can certainly question the accuracy of official police reports in cycling crashes, we certainly aren’t getting more accurate information from the media, since they are themselves getting most of their information from law enforcement.Yes, an insightful analyst can spot inconsistencies or details that show flaws with the reporting, but that rarely tells us what actually happened, it only tells us what may not have happened.
Media reports can only be less accurate than official crash reports, because reporters only report what they’ve been told by law enforcement.
The League illustrates the difference between their methodology and that of official sources:
Approximately 40% of fatalities in our data with reported collision types were rear end collisions. This is higher than what was found in the 2010 FARS* release that included PBCAT-based crash types (27% of fatal crashes with reported collision types), although the crash type “motorist overtaking bicyclist” was the most common collision type in that data as well.
*Fatal Accident Reporting System
Why would overtaking jump from 27% to 40% in 2 years? This speaks more to the problems with their methodology than to what’s actually happening. Their reporting of helmet use by the victim does so as well:
Among the fatalities tracked by Every Bicyclist Counts, only 150 of the 633 reported fatalities included information on whether the cyclist was wearing a helmet. In the majority — 83 or 57% — of those fatalities the cyclist was wearing a helmet. This is higher than other data sources that record helmet use have reported.
Again, this illustrates the bias in their event selection. Crashes involving “responsible cyclists” (sympathetic victims) are more likely to be counted than the ones involving intoxicated homeless persons.
Categorization
The League’s crash typology is nothing like the FHWA crash typing, and vagueness of some of the “types” leave many questions. Without a consistent typology it’s very difficult to compare results. For example “Cyclist side/car front” could mean a number of different crash types. It could mean either a cyclist being hit by a motorist who ran a red light or stop sign, or the opposite. The same goes for a “T-Hit.” Since free crash typing software is available from pedbikeinfo.org it’s a shame that they didn’t take advantage of it so we could compare it to other reports.
“Sideswipe” is normally included in overtaking. The FHWA typing also includes “cyclist sudden swerve,” and “cyclist lost control” in the overtaking group. I cannot tell if cyclist sudden lefts or swerves are included in the overtakings.
[box]
Overtaking crash types:
[tabs slidertype=”simple”] [tab]
Motorist hits a bicyclist who is directly in front of him. Such crashes normally involve a night-time crash with a cyclist without lights or reflectors, or an exceptionally impaired motorist, either intoxicated or asleep.
[/tab] [tab]
Motorist overestimates passing space and clips bicyclist.
[/tab] [tab]
Motorist drifts into bicyclists (usually in shoulder or bike lane). May involve distraction; often occurs on bends or curves.
[/tab] [tab]
Bicyclist swerves into motorist’s path while attempting to avoid a suddenly-appearing hazard.
[/tab] [/tabs]
While a direct hit from behind is what most people fear, that type only makes up a small portion of overtaking crashes. Each crash type has different causal mechanisms and different countermeasures. Lumping them together creates misleading message about the safety of bicycling—specifically, the ability of bicyclists to discourage or prevent motorists from hitting them.
[/box]
Their “unknowns” are exceptionally high, at 30%; probably because media reports tend to be quite sketchy. By comparison, for my review of cyclist fatalities from official crash reports, only 3% fell into the “unknown” category. These usually involved a cyclist found dead after a hit & run. Their categorization also includes “None,” which I presume to be a solo cyclist crash with no other vehicle involved.
The League report’s 40% overtaking was arrived at by ignoring 41% of the crashes.
Already we have 24% not reviewed at all, and another 23% of the remaining 76% set aside as “unknown.” So 41% of the fatalities during the study period were not categorized by the League. The 40% overtaking was arrived at by ignoring 41% of the crashes. The League’s overtaking percentage drops to 31% with the “unknowns” included, which is of course much closer to the 27% overtaking from FARS.
Urban Versus Rural
In the League’s report we are shown that 31% of the crashes were on rural roads and 69% on urban roads (which includes suburban). But their data does not tell us what proportion of the rural or urban crashes involved overtaking. This matters because the League is arguing for separated urban bikeways based on that 40% number.
My review of 105 cyclist fatalities in the Orlando metro area (a predominantly suburban area) since 2006 found 15 overtaking deaths (14%). Of those, nine were rural and six were suburban. None were in an area one would characterize as strictly “urban.”
The State of North Carolina has provided a rich source of statewide bicyclist crash data online using official crash reports and the standard FHWA crash typing scheme. Using their online database I found that 34% of their fatals were overtakings, and 37% of all fatals were urban/suburban, but their online interface would not provide a cross-tabulation of rural/urban and crash type, so we cannot get a direct read on how many urban/suburban fatalities involved overtaking. We’ll have to estimate that from other factors. If one assumes that urban and rural overtakings are just as likely to result in a fatality, then 48 of their 129 fatal overtakings would have been urban, which works out to 12.7% of the 379 cyclist fatalities in their database. But rural and urban crashes are not equal in their severity. The NC data shows rural crashes were 3.6 times more likely to result in a fatality than urban crashes (4.7% versus 1.3%), so let’s divide those presumed 48 urban overtakings by the 3.6 times higher fatality potential of rural crashes. That gives us 13 urban overtakings, which is 9% of their 140 urban fatalities.
While the League report paid legitimate attention to unsafe and illegal motorist behaviors that led to deaths, such as intoxication, distraction and leaving the scene, it’s also a sad fact that many bicyclist deaths involve cyclists behaving similarly. But the League report was strangely silent on this. In North Carolina’s data 19% of fatalities involved an intoxicated cyclist. In metro Orlando it was 40% overall, two of the nine rural overtakings, and three of the six suburban overtakings involved intoxicated cyclists.
Use of lights by cyclists at night is also a key factor. Of the 15 overtakings in the Orlando metro area, four involved unlit cyclists at night; three of the six suburban overtakings involved unlit cyclists. (Cyclist lighting information was not available in the NC data.)
Undermining Lane Control
In his BRAIN interview, Andy Clarke said:
For the longest time it’s been an article of faith that we should be taking the lane, and that separated bike facilities are unnecessary … well, I think we are grown up enough now to say that’s not the case.
Of the report’s five victim profiles, three involved cyclists in bike lanes or on paved shoulders; none were noted as using lane control in a general purpose lane.
And yet their report did not give us any indication of how many lane-controlling cyclists had been hit from behind. Indeed, they didn’t even provide such an event among their five victim profiles. Of the five crashes, three involved cyclists in bike lanes or on paved shoulders; none were noted as using lane control in a general purpose lane.
In the North Carolina data, we can see the differences between overtaking crashes (for all injury levels) in rural and urban areas.
“Bicyclist Swerved” crashes are of course not relevant to lane control. “Misjudged Space” crashes are broadly understood to be remedied by lane control. “Undetected Cyclist” crashes mostly involve unlit cyclists at night. So it is the “Other/Unknown” that would most likely involve the totally distracted motorist who squarely hits the cyclist from behind. 7.5% of these “Other/Unknown” resulted in fatalities in North Carolina, but of course rural crashes are much more likely to result in fatalities than are urban ones.
None of the 105 metro Orlando fatalities involved lane control.
The real “article of faith” being presented here is that a practice promoted by the most experienced and highly trained cyclists in our culture is now being dismissed without evidence.
The real “article of faith” being presented here is that a practice promoted by the most experienced and highly trained cyclists in our culture is now being dismissed without evidence. And it’s being dismissed by a national organization that was originally formed to promote safety and excellence in cycling.
Context Matters
There is no question that we need better reporting of bicyclist crashes by law enforcement, but the League’s report only adds confusion, not clarity. We also should use more than just fatalities when making decisions about street design; adding incapacitating injuries will provide us with far more comprehensive data to combat the most serious crashes.
And no, this review of the League’s fatality report is not intended as a blanket refusal of separated bicycle facilities. There are certainly some roads along which a sidepath would work. High-speed rural highways are good candidates. I used an excellent one in the Bitterroot Valley in Montana. Paths along riverfront and lakefront roads are workable as well. And certain suburban high-speed arterials may be suitable for sidepaths or bike lanes if they have limited and well-managed conflict points. But when it comes to regular urban streets, overtaking is the least of our concerns as cyclists. That is where the turning and crossing conflicts are most common, where savvy, integrated cycling is essential, and where channelization of cyclists leads to unnecessary conflicts.
Vulnerability and risk. Statistics and ethics. Solutions or fixes. Top-down interventions or individual actions. These are the core issues in the long-running bike-lane (or cycle track)-versus-integration argument and in the book Antifragileby Nassim Nicholas Taleb (better known for his previous book, The Black Swan). Antifragile is a long and complex read, but the author managed to summarize it while metaphorically standing on one foot: “Everything gains or loses from volatility. Fragility is what loses from volatility and uncertainty.”
We know what fragile means; the characteristic of being at risk of serious harm or destruction through sudden and extreme shocks. We also have the term robust, which means something is mostly immune to sudden extreme shocks. Taleb noticed there are also things, people, and systems that improve with such shocks, which required the new term, antifragile. There is no hard line between fragile and antifragile; rather it is a continuum. Taleb explores this fragility to antifragility continuum throughout the book.
We bicyclists can’t make ourselves totally antifragile, to the extent that we can benefit from the volatility of traffic, but we can significantly reduce our fragility and become somewhat antifragile by learning to respond effectively to the many small violations we experience. These small violations are often indicators of the actual crash we’re trying to avoid. The close pass indicates the risk for the sideswipe. The right hook close call indicates the potential for a right hook collision. And so on. We cannot and will not become less fragile or more antifragile by putting most of the responsibility for our safety on government or motorists. The best way to achieve it is by changing our own behavior.
Butterflies are very vulnerable, but similar to bicyclists they have the options of movement which reduces their fragility; so much so that they have survived as species for millions of years.
Before we explore how to reduce our fragility as cyclists, it’s important to differentiate between fragility and vulnerability. A china cup is both vulnerable and fragile, we can reduce its vulnerability by protecting it with bubble wrap, but it remains fragile. It’s also useless when wrapped up in the bubble wrap. Butterflies are very vulnerable, but similar to bicyclists they have the options of movement (as well as coloration and chemistry) which reduces their fragility; so much so that they have survived as species for millions of years. (Options is a major theme of Antifragile.)
Bicyclists don’t collide with motorists because they’re vulnerable or slow; but mostly because they’re unpredictable or inconspicuous. (Please don’t misconstrue this as fault; predictability and conspicuity are relative, and of course some motorists are careless. But when we look at the data, it’s clear many — if not most — crashes can be avoided by improving cyclist predictability and conspicuity.) As cyclists, our vulnerability cannot be significantly reduced, but our fragility can. To reduce fragility we must reduce the volatility and uncertainty of our environment, and increase our options.
We must also bear in mind that we’re not in this by ourselves; motorists also need a more predictable environment, because they are also fragile. Yes, motorists are fragile – emotionally and economically. A collision with a cyclist is often damaging to a driver’s psyche or bank account, or both. Well-designed streets and adherence to well-reasoned rules reduce volatility and uncertainty for both cyclists and motorists. As cyclists our options are increased by enhancing (but not violating) the rules of the road with special strategies.
“A complex system, contrary to what people believe, does not require complicated systems and regulations and intricate policies. The simpler, the better. Complications lead to multiplicative chains of unanticipated effects. Because of opacity, an intervention leads to unforeseen consequences, followed by apologies about the “unforeseen” aspect of the consequences, then to another intervention to correct the secondary effects, leading to an explosive series of branching “unforeseen” responses, each one worse than the preceding one.”
—Nassim Nicholas Taleb
Rules Trump Laws (or at least they should)
Adding more rules will not solve the problem, because adding more rules adds points of failure. Instead of more rules, cyclists need more options to avoid the errors of others.
When it comes to rules it’s essential to understand how they relate to laws and to risk. Rules precede laws; they usually arise out of the culture out of experience and consensus. Laws often have political motives and biases; sometimes they uphold the rules, but other times they usurp them. When it comes to risk, particularly in traffic, we must recognize that rules which are violated become points of failure. (This is not always so with laws, because some laws are frivolous or not intended to reduce risk.)
The vast majority of traffic crashes — regardless of mode — happen because one or more parties violated one or more rules. What’s more, even if one party did not violate any rule, they often failed to take effective action to prevent it. I’m reminded of my early auto driving days as a teenager. I was driving along a collector street and another car was approaching from a side street. He was facing a stop sign and I wasn’t. From the passenger seat my dad mildly chastised me for not covering my brake. If that other driver had pulled out and caused a crash he would have clearly been at fault and received the citation, but it would have also been because I failed to take appropriate action. I would have failed to take an option available to me.
Adding more rules will not solve the problem, because adding more rules adds points of failure. Instead of more rules, cyclists need more options to avoid the errors of others. Our most important options are sufficient operating space and adequate line of sight. We can also learn ways to give motorists more options to avoid unnecessary conflicts with us.
So we need a traffic system with the fewest number of rules (not laws) possible, with street designs that support those rules, with the lowest volatility we can manage, and with the greatest options for all. This means we need a system based on what people do best and most naturally, not on what they do poorly or need to be nagged into doing. Many bicycling advocates are calling for systems that increase the failure points and reduce options both for cyclists and motorists.
The rules we have are based on the limits of human perception. Our eyes are at the fronts of our heads and we don’t have x-ray vision. They are also based on years of practical experience, first on open waters where boat captains needed to avoid collisions, and then on roads and streets when cart and carriage drivers needed to do the same. It was only when the industrial revolution came that cities got large enough and traffic got thick enough that the rules needed to be formalized into laws. Eventually traffic control devices were created to help manage the movements and reduce delays. For those who think the rules of the road were created for motor vehicles, note that less than 10% of urban traffic was motorized when William Phelps Eno wrote the first formal traffic codes for New York City in 1909. Eno thought cars were a fad and would be gone in a few years. He wrote his code for horse-drawn vehicles, bicyclists, and pedestrians, with autos, the horse-drawn and bicycles all categorized as vehicles.
The essential fairness of the rules of the road is based on the idea that all people are equal, no matter their mode. That’s why the core rule is:
The basic rules of the road are simple and designed to optimize human capabilities—yield to what’s directly in front of you.
First Come, First Served. You yield to the traffic in front of you because the person ahead of you in the lane was there first. Just like at the grocery store check out. You yield when entering a roadway from a driveway because the people on the road were there first. The same goes when you get to an intersection; you yield to the person who got there before you did. When changing lanes; you yield to the person who is already in that lane. Drivers yield to pedestrians who’ve entered the crosswalk because the pedestrian was there first.
Next comes Keep Right; Pass Left. We have to all agree to keep to the same side of the roadway when going in the same direction, or else it’s chaos. And we want passing to be limited to one side to minimize the amount of scanning we need to do, and because, as I noted earlier, we don’t have x-ray vision. When you pass on the right you can be hidden from view to left-turning drivers approaching from the front.
Intersection Positioning reduces the number of conflicting movements at intersections by eliminating or minimizing converging movements, instead making all movements parallel or diverging.
Yield To the Right at Intersections. If two or more drivers arrive at an intersection at the same time, they yield to the drivers to their right. Why the right? Because motorists sit at the left side of the vehicle and the line of sight is better to the right. (Roundabouts are a special case. They’re really a set of T intersections around a circle, and with the bottom of the T yielding to the top.)
Yield To On-coming Traffic When Turning Left. The purpose of this is to minimize delay and increase certainty. And as noted above, the turning driver needs to be able to see the traffic he’s supposed to yield to.
Notice that with the exception of lane changing, drivers are to yield to people who are in front of them or to the side but still ahead. Our eyes are at the fronts of our heads. Lane changing requires a rearward scan, but lane changing is only supposed to be done mid-block, not at intersections.
In earlier times there was no rule requiring slower traffic to yield to faster traffic. That would conflict with the rule of first come, first served. Today minimum speed laws are in place in many states, but they are only intended to keep drivers of motor vehicles from unnecessarily slowing other motorized traffic. If such laws are applied to bicyclists or other non-motorized users, they are no longer about equality; they are effectively fascist, favoring the powerful over the powerless.
Seeking Order; Gaining Failure Points
Channelizing one type of vehicle on the edge of the road adds complexity, breaks the basic rules of the road and defies human capabilities.
When we add a bike lane or cycle track to a street we effectively add new rules and violate some of the original rules. With such facilities, for bicyclists we now say:
Pass On the Right. The right rear section of a motor vehicle has the worst blind spots, but we expect drivers to check those spots with the same level of effectiveness as when they look forward or into their left side mirror.
Bike lanes discourage proper intersection positioning; cycle tracks often prohibit it. “When do I leave the bike lane to get into the left turn lane? Or should I just turn left directly from the bike lane?”
And for motorists we now say:
When Turning Right, Yield to Through Vehicles Approaching From Your Right Rear (your blind spot), while also scanning ahead for pedestrians in the crosswalk.
When Turning Left, Yield to Vehicles You May Not Be Able to See (because they’re moving toward you while hidden behind other stopped vehicles).
Motorists are often confused as to how to do intersection positioning. “Do I move all the way to the curb as I normally do, or do I stay fully in my current lane?”
So we have added new rules, which means new points of failure. What’s more, these rules are more difficult to obey because they require motorists to look back when they also need to be looking forward, and they require both cyclists and motorists be able to see through stopped vehicles. None of these new rules, or the bikeways that necessitate them, eliminate the need for adherence to any of the original rules, or eliminate the crashes caused by the violations of those rules.
Types of bicyclists and types of bicyclist behaviors.
Bicycle planners are fond of systems that attempt to categorize bicyclists into “types.” The most commonly used type system uses “A” for “experienced” adults, “B” for “novice” adults, and “C” for children. The problem with this system is it assumes certain types of behaviors and preferences for each group, but with no data to support those assumptions. A far better approach is to categorize basic types of behaviors. Some behaviors are more successful at mitigating conflicts and crashes, and some are less successful. Whether a cyclist is a 40 year old man on a $2,000 road bike or a 12 year old girl on a bike from Target is beside the point. Successful behaviors work for everyone. Unsuccessful behaviors fail for everyone. (The presumption that successful bicycle driver behavior is not achievable by novice adults or by children in an appropriate environment is demonstrably false. It’s been done.)
On the other hand, the basic behavior types are driver, edge and pedestrian. Pedestrian behavior is bicycling on a sidewalk, a sidepath, or on the roadway against traffic. Edge behavior is cycling close to the roadway edge with the flow of traffic. Driver behavior is using the roadway in the same manner as any other driver of a slower vehicle, usually in the center of a lane. It’s not a matter of whether a behavior is “right” or “wrong,” but whether it’s successful or unsuccessful for a given type of street and situation. The well-trained cyclist understands the pros and cons for each behavior.
For some strange reason, the opinions of the untrained and less experienced cyclists are held in higher regard than those of the trained and experienced ones. I know of no other activity in which that is the case.
A bicyclist whose default behavior is driver sees a bike lane or cycle track as adding volatility and limiting options. A bicyclist whose default behavior is edge or pedestrian sees a bike lane or cycle track as reducing volatility and increasing options. This is partly because he or she has little or no experience using driver behavior, and partly due to what the majority bicycling culture is saying. It also creates some cognitive dissonance on the part of the edge or pedestrian rider, because he still experiences conflicts while in the bike lane or cycle track.
For some strange reason, the opinions of the untrained and less experienced cyclists are held in higher regard than those of the trained and experienced ones. I know of no other activity in which that is the case.
“What can we say about people who studiously avoid learning from the mistakes of others?” — Taleb
There are better ways to attempt to reduce volatility and improve options that don’t involve violation of the rules of vehicular movement. The best is to reduce motor vehicle speeds in urban and suburban areas, and to limit the number of multi-lane arterials. The faster we travel the poorer our ability to track multiple conflicts and the more distance we need to react to them. Crash data from Florida shows that crash rates for all road users increase significantly when roads are widened from four lanes to six.
We can accept the superiority of bicycle driver behavior over edge and pedestrian behavior by the simple observation that people who effectively learn bicycle driving do not revert to the other behaviors as their default; driving becomes their default because it clearly works better for them (though bicycle drivers will on occasion use edge or pedestrian behavior when it serves a specific purpose.)
“Suckers try to win the argument; non-suckers try to win.”
– Taleb
This bike-lane-versus-integration argument has been running for 30 years at a stalemate, and statistically speaking it’s unwinnable. The argument is in essence: overtaking crashes versus other crashes. The experienced and trained bicycle driver sees the risk of the overtaking motorist as exceptionally low and the risks of turning and crossing conflicts as significantly higher. (Overtaking crashes account for no more than 10% of urban and suburban crashes, while most of the rest involve turning and crossing conflicts.) Based on direct, real world experience they see bikeways that keep cyclists along the edge of the road as increasing their risk for turning and crossing conflicts. Virtually all cyclists who predominantly use driver behavior were once edge or pedestrian riders, so they understand the difference in effectiveness based on real world experience.
Cyclists who predominantly use edge or pedestrian behavior have little to no experience with bicycle driving behavior, so they are unable to compare the effectiveness for themselves.
Advocates of bike lanes, cycle tracks and sidepaths see the risk of overtaking motorists as unacceptably high and believe the turning and crossing conflicts that bikeways create can be mitigated with better design and motorist training and enforcement. They also tie the tenuous safety benefit of bikeways to the desire to get more people on bikes, making the equally tenuous claim that getting more people on bikes will reduce the crash rate (but not the crash numbers). Cyclists who predominantly use edge or pedestrian behavior have little to no experience with bicycle driving behavior, so they are unable to compare the effectiveness for themselves.
If you’re going to trade off one risk for another, you have to first know which risk is greater. Statistically speaking, that means knowing the absolute risk in a measure such as crashes per mile or per hour of exposure. What’s more, you need to know how those risks change with different environments. For example, the risks of turning and crossing conflicts are relatively low on a high-speed rural road with few cross streets and driveways, but one could reasonably expect overtaking crashes to be a potential problem. Conversely, on a low-speed downtown street with many driveways and cross streets, turning and crossing conflicts are a significant concern and overtaking crashes are exceptionally low. Lane control greatly reduces the predominant type of overtaking crash, the sideswipe, in which motorists attempt to squeeze past cyclists within the lane. Therefor, the only type overtaking crash that lane control might not mitigate is the type in which the motorist is so seriously impaired or distracted that he or she does not see the cyclist directly ahead. Such crashes in the urban context are so rare they could be compared to being struck by lightning.
In order to measure the risk for a particular type of crash and how it might change by the type of street, you need to measure quite a few variables, and be confident you’ve measured them accurately. To put this to use, you also need to create a sound mathematical model that reliably predicts the risk for different situations.
Let’s take one type of crash – the right-hook – as an example. In order to predict how many right-hook conflicts per mile or per hour cyclists would expect if they are traveling along the edge of the roadway, we need to know:
How many cyclists travel along the street
How many motorists travel along the street
The numbers of intersections and driveways (minus the ones where there is a dedicated right turn only lane)
The number of signalized intersections
How many motorists make right turns at each intersection
The relative average speeds of the motorists and bicyclists (the closer the relative speeds, the higher the likelihood a cyclist will be in the blind spot of a motorist who’s turning right)
The percentage of motorists who check their right side blind-spots before turning right
And there may be other factors I haven’t thought of. Other turning and crossing conflicts, such as left-crosses and drive-outs, would have their own sets of necessary factors.
Similarly, to measure the absolute risk of overtaking crashes for cyclists who control their lanes we need to know:
How many cyclists travel along the street
How many motorists travel along the street
The relative average speeds of the motorists and bicyclists
The percentage of motorists who are distracted, and how long on-average their attention leaves the road
The percentages of motorists who are intoxicated and sleep-deprived
Also at issue is the relative injury severity of the two (or more) crash types. Once again, overtaking crashes might have high risk of death or serious injury on a rural road, but not on an urban street, while right hooks involving large trucks in low-speed environments have very high fatality risk. These rates are also unknown and difficult to measure.
If you cannot predict which risk is greater, then you cannot ethically restrict a person’s options for limiting their risk based on their own real world experience.
This should explain what I mean when I say the argument is unwinnable, as least from a statistical analysis standpoint. One side can argue that the overtaking risk is higher, but cannot prove it (at least not to anyone looking at the data objectively), and the same is usually true the other way. If you cannot predict which risk is greater, then you cannot ethically restrict a person’s options for limiting their risk based on their own real world experience.
A cyclist who is concerned more about overtaking motorists than about turning or crossing conflicts always has the option of biking along the roadway edge, and often has the option of using a sidewalk. He or she also has the option of controlling the lane, but either chooses not to utilize it or thinks that option is illegal or unsafe. If the lanes on an existing road are narrow and the road has no sidewalks, one could argue that options are limited for this cyclist. He has two options, neither of which he likes. But note that at this point no action has been taken by anyone in authority to change the available options. We could make the lanes wider or add sidewalks; this adds options for this cyclist, and he may or may not prefer these options.
A cyclist who is concerned more about turning and crossing conflicts has the same options as the first cyclist, and will normally choose driver behavior – until a bike lane or cycle track is provided. Now an option has been removed or somewhat restricted. Whether or not there’s a law restricting the bicyclist to the bikeway, the cyclist who avoids the bike lane in order to avoid turning and crossing conflicts gets punished, either socially (through motorist harassment) or legally (through police action). What’s more, the punishment is being doled out to the trained and experienced cyclist who is attempting to avoid a well-document conflict.
A bike lane or a cycle track is a traffic control device. According to the Manual on Uniform Traffic Control Devices:
“The purpose of traffic control devices, as well as the principles for their use, is to promote highway safety and efficiency by providing for the orderly movement of all road users on streets, highways, bikeways, and private roads open to public travel throughout the Nation.
Traffic control devices notify road users of regulations and provide warning and guidance needed for the uniform and efficient operation of all elements of the traffic stream in a manner intended to minimize the occurrences of crashes.”
Notice it doesn’t say “to make people feel comfortable and encourage them to ride bikes.” Nor does it say “support common behaviors” such as bicyclists passing on the right or riding against traffic. Are there any other traffic control devices that expert drivers believe add unnecessarily to volatility, reduce predictability, and eliminate defensive driving options?
There may be areas in which the downsides of a bikeway are very minimal and the upside is significant, such as sufficiently wide bike lanes on high-speed suburban roads, or sidepaths along high-speed rural highways. But a blanket “put in a bike lane” attitude is not a wise strategy.
Laws that eliminate or restrict defensive driving options are even worse. Bicyclists are the only vehicle drivers who (at least in most states) are put in the default position of sharing a lane side-by-side with much larger and faster vehicles. By controlling a lane we get the kind of passing clearance and protection from turning and crossing conflicts that we want, but lane control is framed in our statutes as the exceptional option, not the default option.
How Much Risk Reduction Do You Want?
Would you prefer a 40% reduction or as close to 100% as possible? Published studies comparing streets with and without bikeways vary quite a bit with their claims of crash risk reduction. For the sake of argument, 40% is fair starting point. (I’ll get to the validity of those claims in a bit.) I’ve no doubt you’d rather take the close-to-100% option if it seemed reasonable.
If you go through a Goal Zero process and ignore the most effective strategies for reasons like “people aren’t interested in that” (a common argument against expecting people to seek cycling education), you’re really saying “a 40% reduction is good enough.”
The gold standard of safety programs is the “Goal Zero” approach. With Goal Zero you look at every angle of the system, every point of failure, and strive to understand it as clearly as you can. Then you analyze all the potential strategies to reduce or eliminate those points of failure and determine which strategies are most effective. In bicyclist safety circles that’s called the “3E” approach: engineering, education and enforcement. Enforcement is just trying to ensure people do what they’ve been taught. Too often “education” itself is presented as a countermeasure, but it’s really not. Education is a way to demonstrate to people the types of behaviors that are most effective. It’s the behaviors that are the real crash countermeasures. They don’t work unless the cyclist (or motorist) uses them.
If you go through a Goal Zero process and ignore the most effective strategies for reasons like “people aren’t interested in that” (a common argument against expecting people to seek cycling education), you’re really saying “a 40% reduction is good enough.” Well-designed bicyclist training programs (such as CyclingSavvy) take the Goal Zero approach by addressing each type of conflict and using the most effective countermeasure, not the most popular one.
But I was being charitable to those bikeway studies a few paragraphs back. Their claims of crash rate reductions are highly suspect due to poor methodologies. None of the studies address cyclist or motorist behaviors. They simply look at all the crashes before and after the installation, or compare crashes on streets with and without bikeways. In order for such comparisons to be valid you have to screen out crashes that wouldn’t have been affected by the bikeway presence, such as someone running a red light or a stop sign, or a cyclist entering the roadway from a driveway, or one out at night without lights.
A Canadian study “found” bikeways safer than roadway cycling. This slide demonstrates that authors had no idea what safe roadway behavior is—showing an image of high-risk door-zone cycling and labeling it as “vehicular cycling.”
These studies don’t compare bikeway travel to bicycle driver behavior; they compare bikeway travel to edge and pedestrian behavior without bikeways, because those are the dominant bicycling behaviors. The full range of behavioral strategies has therefore not been studied.
Worse still, some studies have made comparisons between bikeway streets and non-bikeway streets without keeping the other street characteristics consistent. For example, in the Lusk, Furth et al study of Montreal cycle tracks, one-way, one-lane residential streets with cycle tracks were compared to two-way, four-lane commercial streets without bikeways. They also ignored crashes that might have been relevant, but we can’t tell from their paper. (They also omitted a cycle track section that was already known for a high crash rate.) Critiques of that study can be found here and here.
We also have to consider the unintended consequences scenario. If we build a bikeway that is generally believed to be safe, the proponents say, more people will come out to bike. This may well be true, but will those new cyclists keep only to the bikeways? Of course not. And will they seek out training? Not until we get serious about it as a culture. So we will have more cyclists out on all of our streets, but they are doing so with the less successful edge and pedestrian behaviors. The bikeway designs might partly mitigate the failures of edge and pedestrian behavior, but of course not on the streets without the bikeways. So it’s completely predictable that crashes would go up on non-bikeway streets compared to the bikeway streets because we’ve invited a bunch of untrained people into situations they don’t understand, and have reinforced the types of behaviors that get them into trouble to begin with. But the pro-bikeway study authors take this as validation.
The things we don’t measure are often as important, or more important, than the things we do.
“It would be nice if all of the data which sociologists require could be enumerated because then we could run them through IBM machines and draw charts as the economists do. However, not everything that can be counted counts, and not everything that counts can be counted.”
—William Bruce Cameron, “Informal Sociology: A Casual Introduction to Sociological Thinking”
Inhibiting people from learning better strategies is also unethical. How does a novice ever learn to reduce his own risks if training is dismissed as irrelevant or impractical and the infrastructure discourages the more successful behavior?
Bicyclists get hit by motorists while traveling in bikeways. They get rear-ended, right-hooked, and left-crossed. They also get hit by motorists driving out from driveways and cross streets. There are also crashes involving wrong-way cyclists in bike lanes. Yes, of course, these crashes also happen to cyclists who are not in bike lanes, but no-one has solid data showing they are less likely with a bike lane. It’s unethical to limit an individual’s attempts to reduce their risk if you don’t have strong certainty that you have a better strategy. Bicycle drivers know from direct real world experience – which includes all the pertinent factors, just not in the form of quantifiable data – that they have far fewer close calls involving such conflict types when they are controlling a lane than when they are in a bike lane or otherwise along the edge.
Inhibiting people from learning better strategies is also unethical. How does a novice ever learn to reduce his own risks if training is dismissed as irrelevant or impractical and the infrastructure discourages the more successful behavior? Some years ago I met a man who’d spent quite a few years living in The Netherlands. Like so many, he loved their bikeway system and raved about it to me. I didn’t argue with him about it, but just reminded him that here in Orlando you have to know how to handle yourself on roads without bikeways. He assured me he knew how to bike safely in mixed traffic. Each time I’ve seen him since he’s been cycling on a sidewalk.
We could liken bike lanes and cycle tracks to placebos. Even placebos have some positive health effect, but we would never give a placebo when the real drug was much more effective. Vaccines on the other hand work very well. They strengthen the immune system rather than attack the malady directly. Training is a vaccine; it shows the cyclist how to address the most common problems wherever she goes — not just where the government has installed a special treatment — and no-one is misleading her into thinking some other agent will handle her problems for her.
A Robust System
While we can’t make individual cyclists robust or entirely antifragile, we can do it for our overall traffic system. Our traffic system is inherently based on cooperation, but too many people are trying to treat it as though it’s a competitive one. Whether it’s the motorists-as-wild-beasts theme, the media pumping up the us-versus-them angle, or some bicycle advocates with their us-versus-them brand of activism, they are all making the system unbalanced and less effective by bolstering the competitive mentality.
Cooperation is what makes our traffic system robust. It’s what gives it safety redundancy; it’s what makes its users smarter and more effective; it’s what makes traffic flow smoothly; it’s what makes us focus on our own responsibility rather than everyone else’s; it’s what rewards successful behaviors.
Confusion (such as when the system is made too complex) confounds cooperation and feeds competition.
Cooperation is strongest within species and cultures, not between them. So “pedestrians and bicyclists as endangered species” and “bike culture versus car culture” are harmful analogies. We are all one species: homo sapiens. We are genetically and culturally guided to protect the vulnerable of our own species. Protection of the vulnerable is a far more effective driver of human cooperation than the law. A motorist who expects to see bicyclists on all roads (except freeways) and on all parts of the roadway will be more diligent in looking for cyclists than one who expects them to only be on paths or bike lanes.
If we want an antifragile system then, we need three key ingredients: simple rules, cooperation and vulnerability.
Simpler rules increase predictability.
Vulnerability combined with predictability increases cooperation.
And cooperation strengthens the rules.
It’s a virtuous cycle.
“You get pseudo-order when you seek order; you only get a measure of order and control when you embrace randomness.”
Thanks to the tireless volunteer work of John Allen, Robert Seidler, and of course Keri Caffrey, we now can share the first video from the I Am Traffic Colloquium. In this presentation, Keri lays out our take on the core bicycling-related problems we see in our traffic culture, the ill-advised “fixes” that many are promoting, and our own vision for real solutions based on a clear understanding of the problem. 26 minutes of information that will make you really think.
Here is more information on the “I Am Traffic” Bicycle Education Colloquium coming up in February.
The Agenda
Saturday, February 23 9:00 a.m. to 5:00 p.m.
Introductions
Achieving a Vision: Keri Caffrey puts cyclist education and training into the broader context of how communities can genuinely support cyclists
Journeys From Home: Roger & Sharon DiBrito and Robert Seidler share the principles and practices of their elementary and middle school curricula
CyclingSavvy: Keri Caffrey & Mighk Wilson share the origins, principles and implementation of their adult traffic cycling course.
Law Enforcement: Kirby Beck of the International Police Mountain Bike Association discusses strategies to help law enforcement professionals understand the principles of safe and effective bicycle driving
The 6 Es: Dan Gutierrez introduces the goal of developing a comprehensive guide to equitably address the “6 Es” for bicycle driving: Equality, Education, Enforcement, Engineering, Encouragement, and Evaluation.
Sunday, February 24 8:00 a.m. to 2:00 p.m.
Developing a Vision and Mission
Principles: Organizational and Bicycling
Organizational Structure/Business Model
Relationship of Organization to Curricula
Strategic Planning
Leadership
Funding
The Six Es
Accommodations
The host hotel for the colloquium is the Courtyard Marriott Downtown Orlando.
There are a few other downtown hotels within walking distance. Unless you are the “adventurous” type we would not recommend the Travelodge. We also suggest you avoid motels along West Colonial Drive (SR 50) (except the Crowne Plaza) and along South Orange Blossom Trail (US 441).
Meals
There are a few restaurants within a 15-minute walk of the Courtyard Marriott; many more within biking distance. We will give you a longer lunch break both to reach some of these locations and to give more socializing time.
Twice a month since 2006 I’ve taught a course for the local safety council on “alternative” transportation to adults with suspended driver’s licenses. The students routinely recognize the benefits of bicycling when prompted: improved health, reduced environmental impact, reduced transportation costs, increased sociability, and of course, simple fun. To varying degrees they also believe cycling to be slow, dangerous, uncomfortable, physically demanding, and impractical if one needs to carry things. So encouraging people to bicycle is more a matter of removing barriers (practical or psychological) than convincing them of the benefits.
The psychological barriers to bicycling are often more common than the practical ones, so helping people to see the possibilities is a key strategy. People are influenced and encouraged in different ways and through different channels, including the social realm, finances, urban planning and engineering, and the practical aspects of the bicycle and its accessories.
Social
In November of 2012 BicycleRetailer.com writer Ray Keener dug up some data from a 1990 Bicycling Magazine survey. While shops and manufacturers might have wondered at the time if an easier-to-use shifter system might have been the answer to getting more people on bikes, only 12% of “infrequent” cyclists said so. The top motivator the “infrequents” reported was “If I had someone to ride with (46%).” That was followed by:
If I had a more comfortable seat (37%)
If I had a safer place to ride (33%)
If I were in better condition (29%)
If I had a more scenic place to ride (28%).
A good riding partner, or better yet, an entire group of friendly, supportive riding partners can do more to show a new cyclist the possibilities and joys of cycling than anything else. Too often cycling groups have themselves placed limitations on cycling. “We go fast.” “You need a road bike.” “We only ride trails.” “Helmets required.” “Our rides are 30 to 60 miles long.”
I believe the most effective thing experienced cyclists could do to encourage others to ride is simply to ride with them in a place and manner that the newcomer would find comfortable. That may initially mean on a trail or quiet local streets. Don’t set any preconditions. Each ride is an opportunity to share your knowledge and expertise. But take care to open up the floodgates of information. Find just a few things to make improvements on each time — get that saddle at the right height; teach proper mounting, starting, stopping and dismounting; get her generally in the right cadence range but don’t fuss about it; etc. — and spend the rest of your ride simply enjoying the trip. Your task is not to make your friend into an expert rider in a day. It’s to show him that cycling can be enjoyable and that there are lots of things he still needs to learn to make it better. If you agree to give the ride a purpose — to go to a restaurant, a park, another friends place — anyplace they’d routinely go by car — that’s even better.
I’d really like to see each cycling club start a serious new-rider program. Each weekend there should be rides designed for absolute beginners. This doesn’t mean going 15 mph instead of 20. It means taking the one-on-one approach I outlined above and adapting it to a group.
A CommuteOrlando social ride.
In Orlando we’ve been running rides through CommuteOrlando.com and Bike-Walk Central Florida that appeal to those riders. These rides are set up to show new riders the possibilities. Our First Friday rides are social rides of about 10 miles in length that are run all on multi-lane roads. We maintain a well-controlled double pace line at about 10 to 12 mph and controlling the right lane. Ride leaders manage it in such a way that everyone is comfortable and knows exactly what to do. Ice cream rides are held on summer evenings. New riders get to see that there’s no need for exceptional speed or extreme skills to use a bicycle around town. Eliot Landrum in Dallas, Texas has some good ideas on this as well. [link to come]
A woman walks into a marketing and public relations firm and sits down to talk with their lead strategist.
“Our organization has a fun, safe and healthy activity we wish to promote, but we’re struggling to figure out the right approach,” she says.
The strategist thinks for a moment, then responds, “I recommend the approach bicycle advocates have been using for the past 20 years: reinforce the public’s fears about your activity.”
The woman is taken aback, pauses for a moment, then says, “Oh! You had me going there for a moment!”
“What do you mean?” asks the strategist.
“Well, you were joking, right?…”
We must stop being our own worst enemy. While on the one hand we tout the benefits of cycling, too many “advocates” feed the belief that cycling is dangerous.
Finances
Money can be either a motivator or a hindrance in convincing people to bike, or to bike more. For the person of average means, a bicycle can save money, particularly when replacing auto trips. Illustrating the full costs of auto use can be effective with them.
A Get Active Orlando bicycle giveaway for low-income adults.
For some lower income persons an auto is essential, and the extra expense of a good quality bike is beyond their means. Further down the economic ladder, many very low income workers use bicycles, but are limited to poor-quality “big box” bikes. Helping lower income people gain access to good quality bicycles while also providing good cycling education could cast “bicycle advocacy” in a much more positive light.
At the community level, a bicyclist-friendly business program can help in many ways. Businesses that provide discounts or other bonuses to cyclists make people feel good about the businesses and themselves, and such programs tend to interest local businesses rather than chains. Business owners also come to recognize cyclists as customers, rather than just strangers out on the road. Employers and cyclists can also take advantage of federal tax incentives that that allows bicycle commuting costs to be paid for with pre-tax earnings.
A more challenging strategy is removing some of the hidden subsidies for auto use. Parking, for example, is a valuable service that motorists don’t pay for directly in suburban areas (for retail parking all customers pay for parking as the overhead portion of goods and services), and often pay a subsidized less-than-market rate in city centers. Reducing the supply of parking where feasible and getting motorists to pay for parking both out-of-pocket and at its true cost will get people to consider cycling for shorter urban trips (though it really must be done in conjunction with significant improvements to the public transit system).
Planning and Engineering as Encouragement
An inverted U rack; the can’t-miss default for quality bicycle parking.
Of course, along with reducing subsidies for and dominance of auto parking, communities also need to improve the availability and quality of bicycle parking. In a 1995 survey of Orlando area residents, respondents identified good bicycle parking as a more important incentive than trails or bike lanes. Forgotten in the discussion of bikeways and increased bicycle use in New York City is the fact that the City also changed its codes in 2009 to require commercial buildings to accommodate bicycle driving employees. Bicycle theft has long been a serious deterrent to cycling in New York. Leaving a bike locked along the street – even with the best of locks – has always been a risky proposition for cyclists there. It’s quite possible the bike parking code did more to encourage cycling than the bikeways did. Local governments should require quality short-term parking for customers and more secure long-term parking for employees.
Many Americans live in far-flung suburbs, in which important destinations are beyond comfortable cycling distance. Combining bicycling with transit can overcome this challenge. To do so, planners need to provide good connections to and good bicycle parking at transit stops, bicycle accommodations on buses and trains, and other useful amenities, such as bike-sharing systems and shower and locker facilities.
A short path connecting cul de sacs in California.
Trails, bike lanes and other types of separated facilities are often pitched as encouragement as much as they are for safety or mobility. I will leave discussion of the validity of such claims for the Engineering section of this site, but there are engineering treatments that can serve as productive encouragement without compromising the principles of integrated bicycle driving.
Trails in their own rights-of-way and short connector paths can help cyclists stay on neighborhood street networks and avoid busier arterial and collector roads.
Shared lane markings (“sharrows”) and “Bicyclists May Use Full Lane” signs can make it clear to cyclists and motorists alike that bicyclists are normal and expected users of the roadway, and that a centered lane position is usually the preferred position.
Shared Lane Marking or “sharrow.”
Ensuring traffic signal detection systems detect bicycles not only improves safety, but demonstrates that a community cares about the details.
Practicality
Many non-cyclists and inexperienced ones make assumptions about the practical aspects of cycling: that bicycles are inherently uncomfortable; that one can’t carry very much; that bike thieves can’t be foiled; that bicycling requires a high level of athleticism. No doubt many who hold such opinions are simply expressing rationalizations to excuse themselves from having to consider cycling at all. Those people are probably not the “low-hanging fruit” of potential new cyclists.
CommuteOrlando and Bike-Walk Central Florida “S-Cargo” ride.
But others simply haven’t been exposed to the practical solutions to those challenges. On the personal level, we can offer to check that our friends’ bikes fit well and have appropriate saddles. Advocates for cycling can work with their local shops to encourage them to carry accessories like panniers and trailers if they don’t already. Cargo bikes could be an element of a bike-sharing system. CommuteOrlando and Bike-Walk Central Florida lead “S-Cargo” rides, in which the group rides to a local farmers’ market with cargo bikes, trailers and panniers. Those without such accessories also ride along and get to see the practicality of these tools first hand, as do many of the market or store customers and employees.
All of the Above
It should be clear that there is no single silver bullet for encouraging people to bike. Some of these strategies can be advanced by us as individuals; others require organization and sustained effort. Communities that take such an all-of-the-above approach are likely to see real progress.
It’s not Effective Cycling repackaged with a new name.
A common criticism of cyclist education is simply that “it doesn’t work.” Presented with such a statement, I suppose we first have to ask, “work at what?” Those making the claim seem to be saying it doesn’t work at getting more people to ride bikes. I don’t think many are claiming a trained cyclist is just as likely to crash as an untrained one.
Perhaps they may be right that it doesn’t get more people to ride bikes; but on the other hand I know individuals who have most certainly increased their cycling due to education. They have said so themselves. No doubt the “it doesn’t work” claim is based on a belief that few people will take a traffic cycling course.
This debate is at the core of how our society decides to promote cycling.
Do we take an ends-justify-the-means approach in which almost any strategy which encourages people to bike is deemed okay, or do we help each individual maximize their safety, comfort and competence through means which are as ethical as possible? Those who argue for “getting more people on bikes” use cycling as a means to various ends: health, climate change, community livability, etc. Certainly those are all worthy goals, but it has led proponents to take some liberties with science, and misled many people about what factors are important both in increasing the number of cyclists and improving safety. They argue that increasing the numbers of cyclists will make cycling safer in spite of hazards created by some types of bicycle-specific infrastructure.
While it’s true that an increase in the numbers of bicyclists reduces the overall crash rate, the same is true of auto use and walking. Analysis of one study of a Danish bikeway found that the decrease in crash rate was less than would be predicted by Smeed’s Law. To put it in a more direct light: the numberof crashes went upmore than it would have if there had been no bikeway. Increasing numbers of bicyclists canimprove safety, but only given the right circumstances. And even if safety in numbers through facilities does reduce the injury and fatality rates, if those same facilities actually cause some injuries and deaths, then we must question the ethics of such a strategy, especially if other strategies are available which would be less likely to cause harm.
What’s more, proponents assert that the right types of facilities significantly increase cycling. But the scientific support for this is also debatable. There are numerous factors that affect bicycle mode share, including climate, demographics, density, street network connectivity, terrain, the costs of owning and operating a motor vehicle, and most importantly, the presence of a college or university in a community. Do places with bicycle facilities get more cyclists, or do places with more cyclists get more bicycle facilities? It may be both. But if increases in cycling are due in large part to factors other than bikeways, then any reduction in the crash rate is indirectly due to those other factors, not to the bikeways, and if those bikeways cause or contribute to conflicts and crashes — which they do — then providing bikeways as means of increasing use and improving safety does not work and is in fact unethical.
Aside from the debate about whether or not bikeways improve safety, the need for cyclist training is essential. Florida emergency room data shows that two-thirds of adult bicyclist hospital admissions do not involve a collision with a motor vehicle. A significant portion of bicyclist/motorist crashes occur on local streets which will never see any sort of bicyclist-specific accommodation. Crashes involving turning and crossing conflicts occur whether one is cycling on a cycle track, a bike lane, a sidepath, a sidewalk, or on a road with no special accommodation at all. Training cyclists reduces their crash risk for all of these circumstances.
I — and other proponents of bicycle driving — am often assumed to be against all types of bicyclist-oriented infrastructure. That is not at all true. I support and very much enjoy shared use paths that run through independent rights-of-way. Short paths that connect local street networks. Bicycle boulevards. Bike routes and wayfinding that help people stay off of heavily traveled arterials. Shared lane markings on those arterials. Ensuring traffic signal systems detect bicyclists. Well-design traffic calming that keeps motorist speeds down or diverts motorist traffic. Even sidepaths can be a good solution in the right context. None of these treatments (if properly designed) encourages bicyclists to violate the rules for vehicular movement.
A Failure of Marketing, Not a Failure of Education
Another problem with the “it doesn’t work” claim is that it assumes cyclist education and training to be quite limited in definition. Adult cyclist education is presumed to be all the same. For many years, adult education was limited mostly to the League of American Bicyclists’ Traffic Skills 101 course and other descendants of John Forester’s Effective Cycling program. All other adult-oriented courses tend to be lumped together with them.
When detractors say, “We tried education; it doesn’t work (at getting enough people on bikes),” what they are likely seeing is not a failure of a particular curriculum, but a failure to effectively market a curriculum. Marketing is not merely advertising, it is determining what kind of product or service to provide in the first place and how to deliver it to the marketplace. And that, I believe, is where earlier bicyclist education programs have failed the most.
A number of years ago when Keri Caffrey and I were discussing what we’d like to see in adult cyclist education, we were frustrated with the materials we had to work with (the League curriculum) as well as with those who said education was of little value. We knew from direct experience teaching people that education had great potential, but also saw it fail for many individuals. To us that meant the problem was the teaching strategy, not education per se. Sitting in League education meetings and reading posts on the League Cycling Instructor email list, I saw how they were nibbling around the edges trying to incrementally improve a marginally effective curriculum, or arguing endlessly over minutiae such as whether or not to use a mirror. Hardly anyone was asking who their students were, what they want, how they learn best, how they want to manage their time, and how much they might value good training. The League’s curriculum seemed to be based more on what they wanted to teach than on what the customer wanted.
The League’s course is based on the original Effective Cycling curriculum developed by John Forester; many of the long-time League instructors have wanted to return to it. Effective Cycling was originally a 30-hour course, developed in the 1980s. It was intended to make someone into a highly competent sport rider, interested and capable of doing long-distance rides at higher speeds. The League whittled that down into two 10-hour courses, initially named Road I and Road II; later renamed Traffic Skills 101 and 201. While thousands have taken Road I and Traffic Skills 101, only a tiny number have followed up with the second course. TS 101 attempts to achieve the same goal in 10 hours — making an individual into a competent sport rider — as did the original 30-hour course. To do so, students are given a significant amount of information on bicycle components, bike fit, special clothing, minor repairs, hydration and nutrition, cadence and gearing. This leaves relatively little time for learning traffic skills.
For the average person who just wants to bike around his or her community at a comfortable speed, to run errands, visit friends, bike to work, or just have fun, most of those sport-cyclist-oriented topics are wasted time, and make cycling seem more complicated and elite than they want it to be.
CyclingSavvy was designed from the outset for the average adult who wishes to bike around his or her community at a comfortable and sociable speed; who prefers to avoid busy arterials but occasionally needs to use them for short distances to get to a destination or another quiet, low-volume street. Observers of our courses will see very few Lycra shorts or club jerseys, and quite a few bikes with raised handlebars rather than dropped ones.
Core Concepts
Fear was of course an essential issue we had to battle for those intimidated by the thought of cycling amid motor vehicle traffic. That fear is based on beliefs, not on an objective assessment of clearly organized data, so belief-change became a core goal of the curriculum.
As anyone who has waded into some of the hot-button issues of the day can tell you, beliefs don’t get overturned merely by presenting facts. How those facts are presented are at least as important as the facts themselves. So Keri and I did a lot of reading on how to influence people. The most important strategies are getting students to “own” key concepts by discovering the information themselves, and by having them publicly break the taboos of traffic cycling in a peer-supported setting. These strategies are unique to CyclingSavvy.
The other critical strategy was focusing on right-brained learning. Bicycling is a four-dimensional, kinesthetic and social experience, so learning must take place in all of those modes. Few people can fully translate a written or verbal description, or even a static illustration, into a coherent four-dimensional model in their heads. Animation and video get us much further with students. Story-telling in the classroom and the group road tour make it social, and of course on-bike training brings in the essential kinesthetic component.
How CyclingSavvy Differs From Effective Cycling
Since the rules for vehicular movement are nearly universal, especially in North America, the similarities between CyclingSavvy and Effective Cycling are far more prominent than their differences. But the differences do matter. They matter because most people want to minimize the amount of time they spend cycling on high-speed arterials. Depending on where one lives, much of one’s cycling can be accomplished on lower-speed, lower-volume streets. We like cycling on such streets as much for their ambiance as for safety (or the perception of safety). We can ride side-by-side and talk to one another. They tend to be more shaded (important here in the subtropics of Florida). There’s more to see and enjoy.
But sooner or later we need to cross or use a stretch of arterial to get to our destination or to the next network of quiet streets. CyclingSavvy’s strategies show cyclists how to use these arterials in the most stress-free ways possible. Examples include:
Turning right on green even when a right on red is allowed. Doing so gives the cyclist the road to herself for blocks at a time. This can be enough time to get to the cyclist’s target intersection without having to negotiate lane changes with high-speed, high-volume traffic.
Recognizing how traffic flows on the approaches to and exits from intersections. Early lane changes and strategic lane positioning can minimize the interaction and negotiation cyclists must do with motorists while changing lanes.
We also teach a strategy we call Control and Release which is only used for narrow, two-lane streets with significant traffic, or briefly near signals on multi-lane arterials. No-one likes to be the slow vehicle driver with the long line of faster drivers stewing and fuming behind them, but we also don’t like having cars and trucks squeeze past us at unsafe distances and speeds. So we control the lane so motorists don’t try to squeeze past in an unsafe manner, but then, if conditions allow, we move over and strategically allow motorists to pass at lower speed, or at locations where we get some extra width to work with. A similar strategy is also taught for multi-lane arterials when one gets stuck at the front at a red light, and a long queue of cars gets backed up behind us. We simply control the lane going through the intersection with the fresh green, but then pull over into the nearest driveway and let the big platoon of cars go by. After usually just 15 to 30 seconds the road clears out and we have it mostly to ourselves, and we can control a narrow lane without having traffic backed up behind us.
No Need for Speed
A criticism against “vehicular cycling” is that it requires cyclist “go fast” in order to integrate effectively with motorized traffic. While there are some proponents of integrated roadway cycling who hold that belief, Keri and I do not. The strategies we teach work just as well at 10 to 15 mph as they do at 20 to 25 mph. Indeed, a key point we make during the course is that slower speeds confer some key benefits to drivers of any type of vehicle: a more comprehensive view of the environment, better reaction time, and shorter braking distance.
Speed differential is a very over-rated factor. To a motorist driving 45 to 50 mph, it matters very little if a cyclist is going 10 mph or 20 mph. A 30 mph closing speed requires about 200 feet of perception, reaction and braking distance, while a 40 mph closing speed requires about 325 feet, but motorists can easily see cyclists from much farther than that. We’ve illustrated how unimportant this speed differential is with one of our students on a large, high-speed interchange between a high-speed arterial and a freeway. Lateral positioning in the lane is far more important than the relative speeds of the cyclist and motorist.
One need only read the student stories on the CyclingSavvy website to see that speed is of little consequence in getting the average non-sport cyclist to be confident in traffic..
Lane Width and Lane Position
Forester in Effective Cycling and the League of American Bicyclists in their curricula tell cyclists to drive in the right wheel track when the lane is too narrow to share. What we have found, and what has been confirmed by research done by Dan Gutierrez and Brian DeSousa in southern California, is that the right tire track can sometimes be the worst position, particularly on higher-speed, multi-lane arterials. When motorists are driving at 45 mph and more, their decision zone — the range of distance behind the cyclist during which they are best able to decide whether to remain in the lane or change lanes — is quite a ways back due to the speed differential.
The problem is that from that distance, when a cyclist is driving in the right wheel track, it looks as though there is enough lane width for the motorists to pass within the lane. So some motorists will stay in the cyclist’s lane instead of changing lanes at the first opportunity. As they get closer to the cyclist they realize the width is not as great as they thought, but may have now lost the opportunity to change lanes. So they slow down a bit and pass within the lane. This results in a close pass at relatively high speed. Moving towards the center of the lane or left tire track makes it clear from a great distance that the motorist must change lanes to pass safely, so they do so at the earliest opportunity. By doing so they draw attention to the fact that there is something going on ahead, so the next motorist in the lane also sees the cyclist from a distance, and also changes lanes. This is best explained in this video.
Lane positioning is not a simple matter of either keeping right when the lane is wide enough to share or moving into the right wheel track when it is not. In a 9-foot lane the right wheel track is only about 6 feet from the left side of the lane, so the following motorist can see there’s not much width left in which to fit his vehicle. But in a 12-foot lane the right wheel track is about 8 feet from the left side, so many drivers will see that space as adequate. In CyclingSavvy we teach students to focus on how much space is available to his or her left, not to the right.
Changing Lanes
The lane changing instructions in Forester’s Effective Cycling strike us as overly complicated. They are also presented as though wide, sharable lanes are the norm. Lane changing is presented as first moving from the right side of the lane to the left side, then to the right side of the next lane, then to the left side of that new lane. That’s three movements with motorists passing on either side. How many would want to put themselves in such a situation?
What is far more common is for all the lanes to be too narrow to share. So the cyclist is already in the center of the right lane; scans, signals and then moves to the next lane, driving in the center of that one. One move, no lane sharing, very clear.
We also recognize that changing lanes when speeds and volumes are high is daunting and difficult for many people, so we provide a number of left turn strategies that eliminate the need to change lanes and negotiate with high-speed traffic: the “jughandle,” the “box turn,” and yes, even the pedestrian-style dismount-and-use-the-crosswalk (not recommended for Florida and other states where motorists treat pedestrians poorly).
Merges and Diverges
The League system waits until their second course (Traffic Skills 201) to teach strategies for merges and diverges. We include them in the core CyclingSavvy course because we think it is very unlikely that people will take a second course (a position clearly supported by the miniscule numbers of TS 201 courses taught). The fact that our students handle these features with ease during our road tours shows the wisdom of this approach. Interchanges and complex intersections can be enormous barriers to cyclists, so teaching them how to handle them is critical for attaining full mobility in their communities.
Our strategies for these features are very different from the League’s and Effective Cycling. We focus on helping the student read how traffic flows through such features, show them how to minimize converging paths with motorists and help motorists flow around them smoothly. Effective Cycling-based strategies have cyclists negotiate with motorists who are trying to change lanes or pass, often at relatively high speeds, all while sharing lanes. This strategy results in cyclists waiting too long to move into a defensible position and can in some circumstances create more traffic disruption and frustration among motorists. This is very stressful and intimidating for the cyclist.
We show cyclists how to change lanes early so they have control of the lane they wish to be in well before the merging and diverging movements begin. This means motorists are more likely to flow around the cyclist on either side with ample clearance, rather than force negotiation with converging paths. As you can imagine, this is better explained through graphics and video than through text.
While on the road tour, we present students with numerous situations in which leaving the right edge, and even the right lane, is the best strategy. In effect we “wean” cyclists from the curb.
Bikeways
While we do spend a fair amount of time discussing problems that cyclists will encounter when using bike lanes (or cycle tracks or sidepaths), we do this in an objective manner devoid of politics and ideology. The conflicts are real and we simply provide practical solutions. Some of the students themselves, after learning of those conflicts, then question the wisdom of providing such facilities.
John Forester was openly hostile to bike lanes and sidepaths in his curriculum. The League has been fairly neutral in its curriculum, but very supportive of bike lanes in their advocacy. Some League Cycling Instructors have been frustrated with the contradictions between promoting integrated cycling on the one hand, while promoting bike lanes and other separate facilities on the other.
We must be clear though that we do not dismiss all bikeways. We enthusiastically support well-design paths in independent rights-of-way, short connector paths which help connect local street systems, bike boulevards, wayfinding systems, shared lane markings, and other on-road treatments that help cyclists while also supporting integrated roadway cycling. We’d even tolerate bike lanes on some arterials if they provided adequate width and a debris-free surface (but most don’t).
Communication
Rather than expecting motorists to be antagonistic towards cyclists, we strive to create an expectation of cooperation. What motorists want most from cyclists is clear communication on our intentions. We concentrate on communication strategies that assume the motorist is not clear on what we intend to do. (After all, if we watch most cyclists we can see rather quickly why motorists would be uncertain.)
Motorists appreciate clear communication, and going beyond the necessary signals and including ones of appreciation is one of the best forms of advocacy we can imagine.
Skill Building
While teaching Road I and Traffic Skills 101 courses we were frustrated by a number of things. We had only about an hour-and-a-half to teach a handful of skills, and the range of skills went from elementary to advanced. Experienced cyclists were bored until they got to the later skills, and the abrupt climb in difficulty intimidated many novices. Keri Caffrey and Lisa Walker had developed a series of skills for their women’s cycling club, the BOBbies (Babes On Bikes), which was more comprehensive and more progressive in pacing. These skills were incorporated and modified for CyclingSavvy, extending the bike-handling section to three hours. The sequencing of our skills provides interesting, fun skills for more experienced cyclists, while also building progressively enough such that novices are comfortable by the time they get to the more advanced skills.
As with the rest of our course, we strove to make this session fun, and most students report that it is.
Testing
We don’t. Most adults are not interested in meeting some score on a test. They just want to enjoy cycling and feel confident. Testing presents the opposite for many people; it creates an atmosphere of tension. Some people learn well but test poorly. Some test well, but don’t necessarily internalize the content completely.
Instead, we provide an experience out on the streets in real traffic. It is a group tour, but the cyclists drive through with a number of segments on their own. It is not merely a real traffic experience, but also a social experience. Students get to prove for themselves that the strategies we’ve shown them in the classroom actually work, and that experience is reinforced by the other students in the tour. Those who are concerned by an early, intimidating segment can always have an instructor accompany them, but they always volunteer to strike out on their own in later segments, and they all realize the kind of empowerment we’re striving to give them. Those who don’t test well are spared the tension of answering questions and meeting a minimum score, and those who might not have completely bought-in to the concepts have them reinforced by their peers.
Creating a Community
Rather than focusing on club rides which focus on speed and distance, we complement CyclingSavvy with social rides at low speeds around town: ice cream rides, cargo-bike rides to farmers’ markets, and simple social First Friday rides. They are open to cyclists of all skills and speeds. We give them structure and safety, so even those who haven’t taken a cycling course such as ours can see and experience what is possible. They also learn some of the preferred back-street routes and “secret connectors.” Some of those social ride attendees come to CyclingSavvy classes. We are happy to use our area’s trails and connector paths, but prefer to avoid streets with bike lanes as they don’t allow us to ride side-by-side. Bike lanes also present more conflicts for groups than they do for solo cyclists.
Perhaps this strategy — of leading people towards confidence and competence rather than providing facilities which make people feel safer without actually addressing the real conflicts — won’t get as many people on bicycles, or do so as quickly, but we can feel sure that we are supporting our principles rather than subverting them.
Education Is a Gift
“You have such gifts that are important. Just like every species has an important gift to give to an ecosystem, and the extinction of any species hurts everyone. The same is true of each person; that you have a necessary and important gift to give.” — Charles Eisenstein
The nature of a gift is “more for you means more for me.” But look at how bike lanes and cycle tracks affect people. Some non-cyclists see them as taking away from “their” roads. Or they see them as something they were forced to buy (by the government) and give to someone else. Then they encounter them, and the cyclists who use them, and find them confusing and frustrating, especially at intersections. So bike lanes and cycle tracks take the “more for me means less for you” approach. It becomes a turf war. And of course they can only “work” where they are installed, so we “need” to keep taking more and more space from others to get more people on bikes. I spent 15 years trying to prove that bike lanes could increase cycling while improving cyclist safety. In the process I learn they do not. (And other types of bikeways running alongside roadways are even more problematic.)
You cannot convince me something works when every day I see it not working. I see wrong-way cyclists in bike lanes; I get right-hooked by motorists when biking in them myself; I see other cyclists setting themselves up for conflicts. I see the crash reports. While some of these problems are in spite of the bike lanes; many are, to varying degrees, generated by them. And they don’t eliminate many of the real problems cyclists experience. That is not a gift. To anyone.
CyclingSavvy encourages cooperation between road users. Once taught, a cyclist can figure out for herself how to manage any road or intersection. Yes, some motorists get upset seeing cyclists take such an assertive approach, but they eventually realize such cyclists are predictable and cooperative, and that they only need to adhere to the same rules they’ve been following when interacting with other motorists. We’ve also heard from many motorists who find our strategies predictable and cooperative from the outset. So CyclingSavvy is not only a gift to cyclists, but to motorists as well. Those who think cyclists should naturally be antagonistic towards “cagers” might do well to realize that virtually all the new bicyclists we might invite will be coming from the motoring population. How do we convince motorists to become cyclists when they perceive cyclists as adversaries or incompetent?
Those of us who teach CyclingSavvy believe we have a gift we can share. Those we’ve shared it with have thanked us in many ways. That gift is the knowledge that they can hop on a bicycle at any time and go wherever they wish with confidence. By getting on the bike they meet new friends, learn of wonderful new places they didn’t know existed, improve their health, spend quality time with their families, reduce their carbon footprints, and save money. Rather than waiting for the government to provide facilities which don’t work as advertised, they gain all of those benefits as soon as they want them.
You cannot convince me something does not work when every day I see it working. I see people who used to be afraid biking wherever and whenever they wish. I hear their stories of enjoying life by bike. I hear them exclaim how much easier and stress-free it is now to bike. Real, sincere and practical gifts always beat white elephants.
Right up front the League’s methodology for selecting and analyzing crashes has a built-in bias. The report states clearly that:
While the League report paid legitimate attention to unsafe and illegal motorist behaviors that led to deaths, such as intoxication, distraction and leaving the scene, it’s also a sad fact that many bicyclist deaths involve cyclists behaving similarly. But the League report was strangely silent on this. In North Carolina’s data 19% of fatalities involved an intoxicated cyclist. In metro Orlando it was 40% overall, two of the nine rural overtakings, and three of the six suburban overtakings involved intoxicated cyclists.
And no, this review of the League’s fatality report is not intended as a blanket refusal of separated bicycle facilities. There are certainly some roads along which a sidepath would work. High-speed rural highways are good candidates. I used an excellent one in the Bitterroot Valley in Montana. Paths along riverfront and lakefront roads are workable as well. And certain suburban high-speed arterials may be suitable for sidepaths or bike lanes if they have limited and well-managed conflict points. But when it comes to regular urban streets, overtaking is the least of our concerns as cyclists. That is where the turning and crossing conflicts are most common, where savvy, integrated cycling is essential, and where channelization of cyclists leads to unnecessary conflicts.
