« Summer Streets on Kansas Avenue? | Main | Four Mile Run and Custis Trails get paving »


Feed You can follow this conversation by subscribing to the comment feed for this post.

"and remember to slow down or stop to avoid swerving into the vehicular lane if a door appears to be opening."
No! No! No! No!
Dear well-meaning but uninformed naboobs, please go get some cyclecraft. Go take a class from the League of American Bicyclists before making up safety advice which is wrong and dangerous.

If a bike lane is feeding you into the "door zone" where an opening door could impede your travel, the only safe thing to do is to avoid the bike lane entirely. Their advice is like saying:

"When playing russian roulette, be sure not to pull the trigger when a bullet is chambered".

Unless you have X-ray vision, you can't tell if a short person is sitting in that seat, opening the door, until it is too late.

To borrow a quote: "A strange game. The only winning move is not to play."

The left-side bike lanes' main justification seems to be that defective left-side bike lanes are less dangerous than defective right-side bike lanes. While true, this misses the most important point - they are still much more dangerous than well-designed bike lanes! That is the only ethical design engineers should be striving for.

Dave in KY, I agree. And since I'm not in Kentucky, I'll name names: the bike lane on Arlington's Clarendon Boulevard from Highland St to Veitch St is totally unsafe and useless. I use it only to filter when cars are held up. Definitely the worst bike lane on my commute. Washcycle, do the powers-that-be in Arlington know how much this bike lane sucks?

I haven't actually seen anything too bad in the District.

I don't use that one either, especially since it's downhill and I'm usually going as fast, or faster, than traffic. I'm not sure what Arlington County officials think. Maybe Allen Muchnick can enlighten us?

I know a bit of the inside scoop about the left side lanes in Hoboken. They were done as a compromise and maybe not a safe one at that. The roads that have these lanes would be MUCH, MUCH better off with sharrows. The volume of traffic is not high at all and speed limits are at 25mph. These lanes are also on residential streets that DO NOT take you to commercial destinations. The commercial streets in Hoboken that actually need the lanes have no bike amenities because the city government was basically afraid of taking too big of a first step.

I also absolutely detest left side lanes. There are some advantages but it is my professional (yes professional) opinion that those advantages are more than negated by the disadvantages. Along with some of the good reasons brought up already, others include:

1 - They violate a basic rule of the road where slower traffic is to stay to the right and faster traffic to pass on the left. They also reinforce a riding technique that is illegal in many places without the lane (riding on the left side of a one-way is illegal in many places and where it is, it is only legal where one needs to make a left).

2 - That they promote illegal contraflow riding. Many left side bike lanes are ridden in the wrong direction by many riders. (As an aside, left side bike lanes in other countries and in Minnesota are typically designed and intended for contraflow riding. Germany and other European countries come to mind. This technique would seem to violate rules already established in other places).

3 - They expose cyclists to faster and therefore more dangerous motor vehicle traffic traveling in the left lane.

4 - They are often used as a "cop-out" when designers are not allowed to put a bike lane outside of the doorzone. This was the case in Hoboken where designers put the bike lane in the doorzone on the left because the left door is less likely to open. (It could be also argued that this design violated New Jersey's last bicycle facility design manual.)

5 - The driver of a parked car is less likely to be aware of an approaching cyclist as he pulls away from the curb and into the left side bike lane. Drivers have much less situational awareness on the right side and rear of a car and the blind spots are bigger.

To the second half of Andy B's point 1: riding a bike along the left edge of a one-way roadway instead of the right is allowed by the uniform vehicle code and consequently by just about every state in the US.

The point about higher-speed traffic being in the left lane is interesting, though. IMO, in urban riding the fastest traffic on a one-way street tends to opt for middle lanes, where nobody will be slowing to make a right or left turn.

Point taken but then again isn't the UVC going the way of the Dodo?

Still I'll concede that point. Then what about two more reasons why they are bad.

6 - If the street transitions to one-way to two-way operations it is very difficult to transition the bicyclist from the left to the right.

7 - While a passenger side door is less likely to open, the passenger does not have the aid of the mirrors to look to the rear before s/he opens the door into a doorzone left side bike lane.

As for the high speeds, I experience that every time I'm riding in the bike lanes on the Avenues in Manhattan. The left side traffic just over my right shoulder is flying while on the other side of the avenue, traffic will be going much slower.

Plus, who the hell expects a bike lane on the left side?!?! It is so rarely implemented on a national scale.

there are advantages and disadvantages to all bike lane configurations, including providing none.

for every point raised against the left-side design, i can offer a counter-argument against others. should pedestrians cross at-grade or via pedestrian bridges at each intersection? or should we designate special pedestrian streets with no car traffic? or perhaps not allow walking at all? presumably all of these alternatives would result in fewer pedestrian fatalities, but at what sacrifice?

in the end, introducing bicyclists to a roadway that was previously all about cars will never result in zero collisions no matter what configuration you advocate; the objective is to minimize these. to date, there is no conclusive study to demonstrate which design does this best; there is a lot of opinion and speculation (as seen above), there are various precendents, and there are some studies/statistics to demonstrate strengths/weaknesses of various designs. yet no particular design is a clear choice. therefore, it is within the large gray area of engineering judgment that must be exercised by and entrusted to licensed professionals.

andyb, hoboken lanes were not a "compromise", they were based on sound judgment amongst over 10 professionals and legal council, as well as a large community effort after comparing a range of design options against the current traffic patterns and geometry. i respect your difference of opinion, but i think it is careless and misinformed to suggest otherwise.

most important is that despite comments here and elsewhere, there is no legal precedent to knock left-side bike lanes on single-lane, parallel-parked, one way city streets. there is vagueness in the current law, but the spirit of our legal system is such that vagueness cannot be concluded as improper design; rather, due to an evolution in the way streets are used, the law requires updating or, at the moment, interpretation. such updates may be for or against left-side bike lanes, but as of yet this is undetermined. in the meantime, the opinion of legal council was sought and included in the decision-making process. in transportation engineering, this happens all the time, which is why professional licensure is so critical - to confidently stand by the design, with sufficient justification and precedent, if needed.

we can debate all day, and to some extent that is healthy, but without the approriate data, debating opinion (versus fact) is limited in its utility. all that said, if this design is ruled against or demonstrated less prefereable via a study, i would be the first to suggest we make the appropriate changes pronto.

I'm sorry Ian but nobody has begun to come close to proving to me that this is a sound practice. I stand by my reasons why. I can only agree to left side bike lanes in very limited and unique circumstances.

I'm sorry if I'm so adamant about this but as a cyclist with 30 years experience, this idea goes against every fiber of what that experience tells me is safe.

Ian writes:
"to date, there is no conclusive study to demonstrate which design does this best"

There are plenty of studies of crash causes that map it back to cyclist and motorist errors. You can make paint lines all day, but if the paint lines encourage likely crash scenarios, then you're going to get those crashes. If they discourage likely crash modes, then you're going to get safety.

How do left-side bike lanes fare against the common crash scenarios in Cross (1974), Kaplan (1975), and Forester (1983)? These formed the basis for the stats in the League's BikeEd manual:

  • 14% cyclist wrong way riding XXX
  • 11% cyclist left turn from the right side of the road :)
  • 0% cyclist right turn from the left side of the road XXX
  • 5% cyclist swerve in front of car :)
  • 0% Motorist parallel left turn in front of the cyclist XX
  • 11% Motorist parallel right turn in front of the cyclist :)
  • 7% Motorist opens car door into path of the cyclist :)

I did not include some items that don't apply to one way streets, or are clearly symmetric, like motorists running traffic lights.

Items marked with Xs are made worse by LSBLs, when compared to RSBLs.

Items marked with :)s are made better by LSBLs.

Two entries reported 0% in the survey, but would clearly have a higher incidence when cyclists are actually encouraged to ride on the left hand side - imagine for starters that they have the same crash likelyness as their symmetric scenarios, but I'll have more to say on that below.

As I find myself biased against LSBLs, I gave them the benefit of the doubt on the swerving, and accounted a sizable fraction of cyclist swerving to car doors popping open in their path.

The paralell motorist left turn in front of cyclist accrues to the favor of LSBLs because the motorist is closer to the cyclist, and can see more easily, and is more likely to use the left-side mirror because of its greater viewable area.

The only strong bad for LSBLs is the "cyclist riding the wrong way" entry. That is already the #1 cause of car-bike crashes, and its tough to argue that putting the bike lane on the left hand side, no matter how many arrows are put down, will be effective in discouraging contraflow cycling. Heck, contraflow cyclists are all over our RSBLs like flies on s**t already, and they're on the freakin' right side.

Also, it doesn't appear in the study, but there is the likely case of motorist wrong way driving, often by accident, that will place these motorists into head-ons with LSBL cyclists. That's a new danger created, which may or may not turn out to be significant.

To me the contraflow biker case is the breaking point. I think they will cause more causualties than RSBLs. YMMV.

Just wanted to thank Andy for writing his analysis of the problems with left-side bike lanes. It put into words what I was instinctively feeling whenever I saw these (generally in pictures of bike lanes in New York City).

Thanks again.

Thanks to those who agree with me about this. I'm often treated like a pariah on the NYC blogs with my objections to this practice.

I will concede however that the idea to move the bike lanes on the avenues to the left to minimize conflicts with bus traffic MAY (may) be a good enough reason for the practice. This is why NYC started the practice and not for any other. It still leaves bicyclists vulnerable to all the other problems I list.

Sorry, I have a one massive debate a day rule and I've been sucked into the old Purple Line debate lately. I understand the objections to the left side lanes. But I also see some advantages. I'm not saying "let's all switch to lefty lanes now", but the idea merits further study. I do get tired of playing leapfrog with Metro buses.

andy, the "paint lines encourage likely crash scenarios" is a typical "correlation does not equal causation" error in statistics interpretation. in this case, the greater number of bicyclists attracted to bike lanes flips the causation/correlation. to me that neuters this argument.

anyway, i'll bite. perhaps i'm misunderstanding your statistics above, but to me it looks like if you add up the percentages based on the two XXX and :) options, it seems like lsbl's is a design that contributes to reducing significantly more incidents, particularly several incidents with the majority share. what say ye?

The comments to this entry are closed.

Banner design by creativecouchdesigns.com

City Paper's Best Local Bike Blog 2009


 Subscribe in a reader