Research Bolsters Case for Cycle Tracks While AASHTO Updates Guide

For decades, dueling camps of cycling advocates have feuded about how to best accommodate riders. Some have pushed for the construction of Dutch-style cycle tracks, arguing that separated lanes make bicycling safer and less intimidating, while others have insisted such infrastructure isolates riders and makes cycling more dangerous than simply remaining within the flow of traffic.

Why is Montreal outshining every U.S. city on cycle tracks? Photo: ## Layman/RPUS##

Though the debate has grown bitter at times, neither group has had much in the way of rigorous peer-reviewed research to argue their case through the years. However, in the last decade a small but energetic group of academics has started to publish regularly on the topic.

The latest salvo, published online in February and in the current edition of Injury Prevention, comes from Harvard University researcher Anne Lusk. Her study compares crash rates at six cycle tracks in Montreal to nearby streets that had no bicycle facilities, and bolsters the argument that cycle tracks are safer. Lusk found that relative risk of injury was 28 percent lower on cycle tracks compared to the on-street routes.

In addition, she found that about 2.5 times as many cyclists used the cycle tracks than the on-street routes. The finding agrees with the conclusions of a number of other recent studies that show protected bicycle lanes improve safety and help attract new riders.

While cycle tracks are common in European countries, they remain rare in America due to institutional inertia. That inertia was not countered effectively enough by a bicycling movement divided over anti-cycle track arguments made by vehicular cycling advocate John Forester, author of Effective Cycling, in the 1970s and 1980s.

As Jeff Mapes recounts in Pedaling Revolution, Forester helped codify and popularize the idea that cyclists fare best when they are treated as “drivers” of vehicles. He encouraged riders to take the full lane when needed, avoid riding on sidewalks, and move with the flow of traffic.

He also vigorously opposed bike infrastructure, fearing that bike lanes and cycle tracks would give authorities an excuse to ban recreational riders from the road. And he argued cycle tracks and other types of bike infrastructure were more dangerous than on-road riding.

Though a shrill and controversial figure, Forester had an undeniable influence on cycling standards in the United States. He served as the President of the League of the American Wheelman (now the League of American Bicyclists) beginning in 1979, and developed a training program called Effective Cycling that the League adopted.

Within a few years, however, the League warmed to bike infrastructure and he was pushed out. Much to his chagrin, the League started using a slimmed-down version of his training program.

Ninth Avenue cycle track in New York City. Photo: ## DC##

Despite the constant controversy surrounding Forester, one organization that wholeheartedly embraced his ideas was the American Association of State Highway and Transportation Officials. AASHTO’s influential design guide, which serves as the blueprint for most of the nation’s bike infrastructure, echoes many of his arguments about bicycle infrastructure.

The guide, last updated in 1999, recommends strongly against putting separated bike paths near roads for safety reasons, though it doesn’t mention cycle tracks explicitly. And it leaves out many of the more innovative and promising types of infrastructure — such as buffered contraflow lanes — entirely.

This has led, Lusk argues, to the widespread installation of standard bike lanes rather than separated cycle tracks in the United States, an approach that has reduced the number of people — especially women, children, and seniors — who find cycling safe.

Lusk notes, for example, that due to more than 29,000 kilometers of cycle tracks in the Netherlands,  27 percent of trips made in that country are made by bicycle, 55 percent of riders are women, and the injury rate is low. In the United States, in contrast, just one-half of one percent of commuters bike to work, only 24 percent of adult cyclists are women, and the injury rate is at least 26 times that of the Netherlands.

AASHTO began circulating an updated draft of its guide in February for comments. Unfortunately, the updated draft, though somewhat more inclusive, still leaves out critical life-saving types of infrastructure — notably cycle tracks and buffered bike lanes.

According to Fairfax Advocates for Better Bicycling, AASHTO has already received more than 1,500 comments on the revised guide and hopes to produce a revised draft by May or June.

Leaving cycle tracks out of the guide would alienate most cycling advocates. Some American cities have already started installing cycle tracks, and an alternative guide produced by the National Association of City Transportation Officials includes many of the same promising types of infrastructure that AASHTO continues to omit.

Though AASHTO has taken some positive steps to create national bicycle routes, the group recently raised the ire of cyclists for arguing that state DOTs shouldn’t be required to adhere to a new federal policy that puts pedestrians and cyclists on a more equal footing with motorists. They’ve reversed that policy. Here’s hoping they reverse their aversion to cycle tracks as well.

52 thoughts on Research Bolsters Case for Cycle Tracks While AASHTO Updates Guide

  1. Although undoubtedly well meaning, John Forester has kept the US decades behind in bicycle infrastructure and as a result mode share and safety.

  2. The question I always ask appointed cycling advocates and DOT officials: “would you recommend that your mother, wife and/or children cycle on these 8 lane roads with STR signs that you define as safe”. I usually get a blank look or the answer “of course not”. Then I get the long explanation that this are the accepted-recommended standards.

  3. Not getting out of bed in the morning is safer, riding my stationary bike is safer. What’s the point? You still have to get out in traffic on a bike to get where you’re going in this town.

    BTW: Are Cycle Tracks in the Master Plan?

  4. A couple of comments.

    While totally disagree with John Forester about being total against bicycle infrastructure, his vehicular cycling does work! What do you do when the bicycle path, lane, cycletrack ends?!?! Vehicular cycling is the answer and all cyclists can benefit from understanding its basic principles.

    I love Anne Lusk but one could say that she is equally biased in favor of cycletracks as John Forester is opposed to them. She is a great advocate for improved and separated bicycle infrastructure. That said, I would rather see a report proving the claimed safety of cycletracks coming from someone without a track record of advocating for them.

    There was also good reason for bicycle advocates to opposes separated bicycle infrastructure in that past and that was because so much of it was horribly designed and many states and other jurisdictions still have laws that require bicyclists to use them.

    European cycletrack designs often leave much to be desired in my opinion so don’t think they’ve got it solved either. I think that NYC has done a decent job of bringing the cycletrack to the next level but they still haven’t perfected it yet even though in some places they have come damned close and in others I still cringe!

  5. Can they be designed to allow paratransit vehicles to load passengers at the curb as required under ADA?

  6. John Forrester’s ideas might have worked, if not for the invention of SUVs, road rage, and distracted drivers.

  7. In some cities, the cycle paths do not end. Yes there are techniques to know when you are caught cycling in traffic, but my preference is always towards properly build, segregated infrastructure. I like this article because it points out the damage these self-righteous “vehicular cyclist” assholes have been causing – and it’s time they were brought to book.

  8. Yes, vehicular cycling works- it’s a method of survival. It does not make streets any safer or appealing for cyclists of all ages, as segregated bikeways do.

    Also, each country in Europe has it’s own traffic laws and guidelines for planning and priorities in road building. One must also remember that what exists now has been planned and built maybe decades ago, copying the ideas from bikeways that other nations had built before that. They might get built differently today.

    I think the best bikeways can be found in the Netherlands. Do not fall for cheap fakes from other places claiming to be bicycle friendly. See David Hembrow’s blog ( ) where he explains the Dutch bikepaths, and shows examples good of intersections and traffic circles. He even commented the american standards:

  9. And Vancouver and Paris and Copenhagen and Barcelona and Sevilla and… ON AND ON AND ON. Biking will never truly work for the mainstream in the US because we’re assholes.

  10. John Forrester’s problem is not his ideas, but that he’s an argumentative jerk, and a pied piper for an army of argumentative jerks.

  11. The real problem is that there is a difference between a cycletrack and a widened sidewalk. In Much of the US thats exactly what they are building; 8-foot sidewalks with no engineering at the intersections. That is stupid and dangerous. I welcome seperated facilities so long as they are engineered properly to mitigate the intersection danger.

    There is a vast middle ground between Forester and the avg cyclist who likes the idea of being “seperate” but has no knowledge of infrastructure.

    I wish we would stop referring to things as “seperated” because y

  12. Of course we should disregard all research and the experience of other countries in implementing bicycle infrastructure! We should especially disregard the example of the Netherlands! 26 times lower injury rate than the US–how pathetic is that? Such a wimpy statistic should be ignored like the puny little number it is. Unless the injury rate is *26,000* times lower than the US it’s ridiculous to call for the end of vehicular cycling. And anyway, current cyclists are perfectly happy with the US injury rate. It’s all the wimpy non-cyclists who seem to think the rate should be lower before they deign to get on a bike.

    And then there’s the 27% of all trips in the Netherlands made by bicycle! As if I’m impressed. As if this is some kind of sociological revolution. Get back to me when it’s 95% of all trips, then we can talk.

    Until then, we will be proud of our 1/2 of 1 percent commuters. This is so close to 27% that I can’t believe it’s even mentioned. It’s laughable that some would argue this shows the value of separated bicycle infrastructure. Since vehicular cycling is such a success story on every level, it’s ridiculous to even consider other approaches. After all, I once saw a really bad bike lane put in, and this means every bike lane America will be just as bad! We can’t possibly go to Netherlands and study what they’ve done or import some of their urban engineers to help us do it right. This is harder than rocket science and brain surgery all rolled into one!!!! All change will be worse than what we have. Doing nothing is the best path.

    After all, there is no urgency on this issue. It’s okay to have congestion and pollution and CO2 emissions for the next two or three decades. In the end, we’ll be glad we were such sluggards because it’s far better to have people die from being run over, diabetes, lung pollution, and climate change, than to even slightly inconvenience one current vehicular-loving cyclist.

  13. Here’s a video from the Netherlands with a camera equipped bicycle traveling down a cycle track. Some arguments against cycle tracks are given in a blue box in the top left of the screen and counter arguments for them are shown in a green box at the bottom.

    Here’s another video comparing American lane intersection design guidlines to how the Netherlands approaches the problem.

    Another video example describing how the Netherlands designs intersections for bicycles compared to Holland, Germany and new bikeway designs in the city of London..

    A video showing several examples of how the problem of bus stops interfering with bike lanes in the Netherlands city of Utrecht are dealt with.

    Another Netherlands video showing how a busy driveway for cars crossing a cycle path can be designed for safety.

  14. The problem with the dutch path network is it’s designed for short-haul trips, typically 1 mile or less. If I want to ride longer I don’t want to be on separated paths.

  15. The problem with the dutch path network is it’s designed for short-haul trips, typically 1 mile or less. If I want to ride longer I don’t want to be on separated paths.

  16. The problem with the dutch path network is it’s designed for short-haul trips, typically 1 mile or less. If I want to ride longer I don’t want to be on separated paths.

  17. The problem with the dutch path network is it’s designed for short-haul trips, typically 1 mile or less. If I want to ride longer I don’t want to be on separated paths.

  18. Why not? I’ve ridden on cycle-tracks from Copenhagen to Taastrup, Denmark to get to a shop. 20km each way. It was lovely!

  19. Dennis,

    I’ve seen the Mark & Lei videos on YouTube and they are very compelling and demonstrate the cutting edge on Dutch Cycletrack design. Much of what is shown is very well designed and American designers could learn a lot from what is shown. I would have no riding on those facilities but then again the Dutch have been working on incorporating bicycles into their transportation landscape for a long time.

    What I’ve seen in Germany, that I travel to often to visit family, is more of a push towards slow-speed, shared use streets in the cities and towns where the basics of vehicular cycling comes into play (staying to the right, signaling, obeying traffic signals, etc.) Outside of town, there are field and forest roads that do an excellent job of providing alternative traffic free (very light) routes between towns. Elsewhere outside of town sidepaths are used next to the high trafficked roadways but there again there are very few cross streets or driveways if any at all.

    And that’s the problem with most places in the US that makes the use of cycletracks all but impossible, the high density of driveways per mile. Imagine the turning conflicts with a cycletrack on a San Fransisco street of single family row homes where every home has a driveway. Or on a state highway out in the sprawl where traffic would need to constantly turn off and on the highspeed highway, across the cycletrack. I don’t know if there is a cycletrack solution to those design problems that wouldn’t require a huge amount of money to totally redesign everything!

  20. Excellent point Robert. Over here in California, the only difference between a sidewalk and a “bike path” is about 4 feet of width. Both require using uncomfortable buttons at intersections (while the road gets sensors) and placing straight traffic to the right of right turning traffic.

  21. I live in Montreal, and I was hit directly from behind while “taking the lane” on a street with no bicycle facilities. There is absolutely no way the guy didn’t see me, and I suspect he was speeding to pass another car and distracted. Luckily I survived. The point is, vehicular cycling is only as good as the worst driver around you, just as cycle track cycling only works as far as the extent of cycle tracks.

    In a perfect world, everyone would drive courteously and there would be no need for cycletracks. In the real world this is unlikely. Getting everyone to drive safely is very difficult, but installing quality cycle tracks is relatively easy.

  22. I find an argument in favor of cycle tracks that is frequently hinted at, but isn’t explicitly called out in these types conversations (see the blog post above), is that the perception of safety cycle tracks afford (weather real or perceived) leads to a positive feedback loop that does indeed improve safety along corridors where cycle tracks have been installed.

    The argument reads like this (the bullets below contain links to random research backing up the statements above. These were based on a cursory 30 second Google search and should not be considered a comprehensive lit review):

    1. People feel safer biking in cycle tracks (especially women and children).

    2. Cycle track goes in, bike mode share goes up in the corridor.

    3. Presence of increased cyclists as well as cycling infrastructure in the corridor changes driver behavior.
    Basically, the cycle tracks’ visual prominence in the right-of-way as well as the increased number of bikers using the street all serve to alert the drivers using the street to expect bikes. Drivers pay more attention and drive more cautiously.

    Accident rates go down

    So… As to weather or not the increased safety is caused by perceptions of safety or an inherent improvement in the design…who cares.

    Accident rates have been reduced, bike mode share went up, and…extra bonus, traffic was calmed. Win win win.

    AASHTO’s insistence pushing Forester’s a meme that cycle tracks don’t work do American road users a disservice.

    And…given the research showing that women are much more likely to chose to travel by bike after bike infrastructure like cycle tracks (and bike lanes too for that matter) have been installed…And are thus most likely to be negatively impacted by AASHTO’s destructive policy…

    AASHTO’s insistence pushing Forester’s a meme carries an odorous whiff of misogyny.

  23. Vehicular cycling works better in suburban and exurban sprawl where friction from side streets and driveways is limited. But that is not to say its an acceptable cycling environment for women, children, or the elderly.

  24. Traffic laws and driver training are very different between the US and Netherlands, as are typical cycling speeds. These differences somewhat reduce the inherent dangers of sidepath/cycle track cycling in the Netherlands, but make sidepaths/cycle tracks more dangerous than normal roadway cycling in the US. Right-on-red prohibitions and much stronger enforcement of crosswalks laws are examples.

  25. Because it is optimized to minimize junction hazards, I find bicycle driving in travel lanes according to the rules for drivers of vehicles is even more important in downtown urban areas with lots of intersections and driveways than on suburban/exurban arterials with fewer junctions. In an exurban area with just one or two junctions per mile, a sidepath or cycle track will have fewer hazards. In a downtown, however, unless it’s a waterfront path or rail trail the sidewalk or sidepath will feature multiple hazards per block. If urban sidepaths are built, it’s especially important that the law and corridor designs treat them as optional for cyclists.

  26. I’ve cycled from Aachen to Maastricht and back recently. The whole trip was 45 miles in total and it was on separate cycle tracks all the way. I didn’t have any problems being isolated from the car traffic and in fact the presense of cycle tracks was the reason I took this journey in the first place.

  27. I think that the Lusk paper adds little to the conversation. In short, if you compare a six-lane two-way traffic commercial boulevard to a one-way residential street with parking on both sides, one should not be surprised that the collision rate is different.

    Fundamentally, collision data is pretty sparse such that estimating the tiny risks associated with cycling is highly dependent on the mental model that people — either explicitly or implicitly — use to interpret the data. A lot of people fail to recognize this, IMO.

    Regarding bicycle accommodations in general and the context of the article, I think that one can build facilities where the increase in risk in an all-things-equal-context is pretty small and transportation efficiency is still quite good. But cycletracks — particularly contra flow — in areas with lots of intersections is not one of them.

  28. In rural Japan they have cycle tracks along many if the train routes, and I’ve traveled pretty far on them–25+ miles? They’re often very direct routes fbetween population centers with shallow grades. Way shorter and flatter than regular roads.

  29. You’re points are well taken, but you didn’t even mention the role angels and freedom play in everything that happens in America, unlike in Europe or Canada. Oppressed, godless communists can try bike lanes, but to suggest that we could learn anything from their experiences is absurd.

  30. You don’t have a case for bicycle infrastructure, bicycle infrastructure is the roadway. If you want special facilities you need to have bicycles declared non-vehicles and remove their entitlement to the roadway.

  31. John Forester has deliberately worked to keep the U.S. decades behind because he objected to having to share the road with “slower” cyclists. With no infrastructure in place, the roads were kept clear of the “inexperienced” cyclists that were “in his way.”

  32. I’m afraid that’s wrong. Dutch cycling infrastructure is as continuous and comprehensive as the national road network, if not more so. I have ridden on it as a long distance touring cyclist and ever since I have been 100% in favour of bringing it to the UK.

    I am lucky to have similar infrastructure near my home in Watford (although by Dutch standards it is laughable). It makes my ride to work/the shops/the train station/the pub an absolute pleasure.

  33. As a female bike commuter of the street this article is about, I find this super offensive. There are tons of different types of people already of different ages and genders cycling on this street every day. It’s not a street people fear to bike on [other than maybe the bumpy surface and lack of street lighting at night]. You’ll get more people out on bikes if you just extend and widen the bike lanes and repave the surface. The actual design does not extend the entire length of the street. There’s also the ability for cars to park on the side without parking and delivery trucks and taxi cabs and people waiting for passengers are for sure going to take advantage of it. So you really only have protection on one side of the street with the cycle tracks, and normal street level bike lanes on the rest of the road. This cycle track design for this street is just bad. Not all cycle tracks. That’s the major point this blogger decided to miss when criticizing those who disagree.

  34. No money, no infrastructure, no cyclists, no “right to the road”. Rights only exist if we are comfortable and safe in using them. And you will NOT be safe on my commute. Three of my friends who are all experienced VC, one is an instructor, all got hit by cars. We regularly grind up VC and spit them out!

    I ride on the sidewalk and the shoulder. Never been hit!

  35. Voiland’s article demonstrates the strange superstitions prevalent
    among bicycle advocates.

    Voiland claims that Forester (that’s me) “fear[ed] that bike lanes
    and cycle tracks would give authorities an excuse to ban
    recreational riders from the road.” Recreational cyclists? Voiland’s
    claim states that he expects the police to stop cyclists along the
    road to demand their traveling purpose. “Are you going to work? Then
    OK.” … “Are you out for enjoyment? Then get off the road.” Voiland
    probably is not smart enough to have thought that far ahead; he is
    expressing only the unthinking antagonism felt by bicycle advocates,
    who think only of useful cycling, against vehicular cyclists, who
    like cycling so much that they do large amounts of both recreational
    and transportational cycling.

    Voiland’s implication that fear of being banned from the roads was
    foolish demonstrates the ignorance typical of bicycle advocates. The
    American AASHTO bikeway system was designed in California by
    government-established committees, the California Statewide Bicycle
    Committee and the California Bicycle Facilities Committee, both
    operated by governmental highway and motoring officials with only
    one cyclist permitted on each of them, me on the first committee,
    John Finley Scott on the second. It was clear that the object of the
    committees was to restrict cycling to only the edge of the roadway,
    or off it wherever government built side paths (now called cycle
    tracks). The fear then was real; only the efforts of vehicular
    cyclists, demonstrating that the proposed restrictions were so
    dangerous for cyclists that the dangers jeopardized motordom’s plan,
    got the laws and the designs modified to make them less dangerous.
    That history has been published for forty years, but bicycle
    advocates refuse to learn it.

    Voiland is correct that the AASHTO bikeway designs followed the
    California designs, but he gives the wrong reason. Both AASHTO and
    the USDOT adopted the California designs (and the Uniform Vehicle
    Code adopted the California laws), while abandoning the USDOT
    designs already published, because they accepted the idea that, with
    the most dangerous ideas winnowed out by the intervention of
    California’s cyclists, these presented the most restriction that
    could be imposed on cyclists without running the risk of losing
    personal injury suits brought by dead or injured cyclists.

    There’s no mystery about this; the history has been told for
    thirty-five years, without contradiction. Yet the bicycle advocates
    refuse to accept it. One would think that bicycle advocates would
    hurry to join in the opposition to motordom’s policy of restricting
    cyclists, but they don’t. They don’t because, although the American
    bikeway program was designed by motorists to restrict cyclists, its
    popularity overwhelmed bicycle advocates’ anti-motoring views. For
    seventy years, motordom has pursued the policy of removing
    inconveniences to motorists by, in the case of cyclists, frightening
    them to the edge of the roadway, or off the roadway if possible.
    This is done by creating fear of same-direction motor traffic, fear
    so great that it overwhelms all the more important aspects of
    cyclist safety. Therefore, bicycle advocates, such as Voiland,
    cannot believe historical fact because that fact contradicts their
    faith that protection from same-direction motor traffic is the most
    powerful means of making cycling safe. There it is, fact versus

    The Montreal cycle-path study by Lusk et al shows the defects of the
    same faithful ideology. It compares car-bike collision rates on
    supposedly similar streets, some with cycle tracks, some without.
    The ideological faith in bikeways as anti-motoring devices motivates
    people such as Lusk to undertake such studies. Because these people
    have very little knowledge of traffic operations, they were unable
    to observe the differences in traffic conditions of different
    streets. In the discussions by people with bicycle traffic
    engineering knowledge, held after the publication of Lusk’s paper
    (discussions which should have been held by the referees of the
    paper, had they had the qualifications and knowledge to do so), it
    has been conclusively demonstrated that the supposedly similar
    streets do not have similar traffic conditions. With these errors
    corrected, it is likely that the data of the study demonstrate that
    the cycle tracks increased the car-bike collision rate.

Leave a Reply

Your email address will not be published. Required fields are marked *


Cycling Up 70 Percent on London's Bike Superhighways

Figures are starting to come in from London’s investment in “cycle superhighways,” and the data make a strong argument for more robust dedicated bike infrastructure. A Transport for London study found that cycling is up 70 percent along routes where the city’s beefed-up bikeways were installed, reports Network blog Part of the increase is […]

Cycling Up 70 Percent on London’s Bike Superhighways

Figures are starting to come in from London’s investment in “cycle superhighways,” and the data make a strong argument for more robust dedicated bike infrastructure. A Transport for London study found that cycling is up 70 percent along routes where the city’s beefed-up bikeways were installed, reports Network blog Part of the increase is […]

The Next Breakthrough for American Bike Lanes: Protected Intersections

As protected bike lanes become more widespread in the United States, creating physical separation from motor vehicle traffic that makes more people comfortable cycling on city streets, advocates are starting to push for even safer bikeway designs. One area where the current generation of American protected bike lanes leaves something to be desired is intersections. […]

Complete Streets Planning Becomes Law in Hawaii

In more and more communities around the country, the benefits of complete streets — designed for the benefit and safety of all users, not just automobiles — are becoming clear. The latest advance comes in Hawaii, where the governor has signed legislation that makes building complete streets a state policy. Today on the Streetsblog Network, […]