“Sharrow” striping is one of the stupidest ideas ever devised by road planners. Designed to avoid political conflicts on removing parking or auto lanes — they are quite useless for cyclists. Indeed, sharrows may increase the risk to cyclists according to a recent UBC study. Lots of good data in the study, but here is the main result:
The study also found traffic circles to be extremely dangerous. That is a surprising and controversial result.
Update: Here is the main link to the study group, with some additional info.
THANK. YOU.
It looks like Sharrows were more unhelpful on high-traffic streets, which is where they are “supposed” to be used. I’m not surprised by the results of the study. Basically, the sharrows are marking high-risk areas, where heavy car and bike traffic are sharing space on a busy street; this is where we would expect to see higher collision rates.
A better study would compare collision and injury rates before and after sharrows were installed; I’m not sure if we can conclude that sharrows actually increased danger, based on this study.
Here in Portland, sharrows have been placed on low-traffic “Bike Boulevards” (like the side-street bike routes in Berkeley) to help with wayfinding for bikes, and to remind drivers to expect bikes. This seems to work well, though it’s not what they were “intended” for.
I agree with Joseph, you cant draw a conclusion from this, especially from just slides.
Note also according to this study, theres less risk when traffic speeds are highest than at slightly lower speed. Id wager it’s because cyclist avoid the highest speed streets. So raising speeds does NOT make cycling safer, although if you draw your conclusion the same way, thats what youll get.
Id say what happened here was:
-Route is dangerous
-Route gets sharrows
-Route gets safer, but still more dangerous than route with bike lane or cycle track.
You know what the conclusion reminds me of? The crap we here about crosswalks and how they endanger pedestrians. Same type of false conclusion was drawn.
Lets see real data showing the street before and after sharrow, and before and after traffic circle, not comparing apples to pomegranates.
And I completely disagree that sharrows are useless. Until we force drivers to do the written test every renewal, sadly signs are the only way to communicate new laws or misunderstanding in laws. Thats why those little signs that say “stop for pedestrian in crosswalk” work so well, because of all the idiots that dont know the law.
Sharrows are the same type of education for the “get on the sidewalk!” crowd.
Spot on, JJJ. There are many examples where sharrows have had a huge impact on safety. Every road and every treatment is unique. Trying to make this a one or two size fits all analysis is an epic fail.
I agree, let’s see the data before and after the install. Also, they should talk to folks in Ithaca, NY and find out how they like their sharrows. They’ve had them on city streets for almost 10 years now. I rode there a few times and felt built into the system, and the drivers were used to and expected bikes to be in the road.
The sharrows in Ithaca are quite effective. Note several things:
(1) The sharrows are in the middle of the lane, where they need to be
(2) The street width contains enough room for one traffic lane in each direction and one parking lane in each direction — and that’s all.
(3) Car speeds are slow.
The sharrows act as an instruction to drivers to *not* pass bikes on the left with no space at all, which is (sadly) what drivers tended to do without the sharrows. That’s what they’re good for.
I would never ever use them on a fast-moving road, and if you don’t put the arrow in the center of the lane, you’ve missed the entire point of sharrows.
To clarify an issue raised by later comments, the streets with sharrows in Ithaca are *busy* but not *fast*.
To answer a further implied question, removing the parking is not an option on many of these streets, because most of the parking is truck loading zones, handicapped parking, fire lanes, etc. etc.
Those are bad sharrows in the image you’ve posted. They need to be in the middle of the lane or even a little to the left of that so that cyclists and motorists can see that cyclists have a right to the whole width of the lane.
Putting them in the door zone is bad practice.
And traffic circles being dangerous is not astonishing to anyone who has even had to ride through one. They are known to increase motorcar collisions, too, and they’re bad for pedestrians.
Traffic circles are terrible for every purpose except moving more motor traffic faster and that should be the last priority of urban street engineering.
We have a few traffic circles (roundabouts) being used for their correct purpose here in Ithaca. The correct purpose is when you have an intersection which already exists, and which you are stuck with because of where the buildings have been built or where the topography is — where a whole bunch of roads (five or more) come in at a single point at completely random angles to one another. This is also why you have so many roundabouts in the UK, with ancient roads.
There simply isn’t another traffic control device which *works* for those intersections. I’ve tried to use ones equipped with lights, ones equipped with stop signs, ones equipped with yields, and it’s a mess, confusion, danger, and crashes. The roundabouts work reasonably well.
Important to note: those intersections were dangerous for pedestrians *before* the roundabouts were installed. The roundabouts don’t make them any *worse*. The roundabout traffic is also very slow-moving.
If you’re laying out a street grid, you wouldn’t want to lay it out that way, of course. But if you’ve already got something like that going on, a roundabout makes sense.
This is a very interesting study, thanks for covering it.
I read through the presentation and didn’t see any info about the severity of the injuries involved in the individual collisions. Since lower traffic speeds have been shown in other studies to result in less severe injuries it makes me wonder if the choice is between traffic calming with more, less-severe injury collisions or no traffic calming with fewer, more-severe injury collisions. Also, I have a difficult time understanding how speed humps or sidewalk bulb-outs might result in non-intersection collisions, so perhaps this category needs to be expanded out further for a more distinct referendum on different traffic calming tools.
As for sharrows, I have mixed feelings. On multi-lane streets I agree that they are almost always a bad idea, but for narrow neighborhood streets I have found them to work well as supplementary wayfinding guidance, especially on bike routes with lots of turns. Also, they can be used to direct cyclists onto already bike-friendly corridors, and by condensing the bike ridership onto specific streets the safety in numbers rule applies. I can also report anecdotally that cyclists in my neighborhood tend to ride further away from the door zone on narrow streets with sharrows, striped correctly toward the middle of the lane.
I’ve never really liked sharrows either. They’ve always seemed like a political cop-out when used as “traffic control”. Not really helping cyclists and confusing motorists. Some ppl. think they’re end-all-be-all of infrastructure bridging the gap between VC’s and those wanting separated infrastructure. I do agree they make nice route markers on quiet side streets, but that’s about it.
Nice post.
[…] City looks at how climate change is impacting cities around the world, beyond Hurricane Sandy. And Systemic Failure shares research that casts doubt on the value of the sharrows, even showing they can make streets […]
[…] City looks at how climate change is affecting cities around the world, beyond Hurricane Sandy. And Systemic Failure shares research that casts doubt on the value of sharrows, even showing they can make streets less […]
Besides the flaws in logic/methodology pointed out by others, you are showing a single slide out of context, while skipping the one that shows that huge error bar for the sharrow treatement. The bottom of the error bar is actually in the “sharrows are safer than nothing” territory, and even overlaps the “painted bike lane” error bar.
I think it’s to early to conclude anything, but I do think that sharrows can be useful in some places, mostly as an educational tool both for drivers and for cyclists. You can put them on narrow roads where there’s no room for a bike lane, to tell the drivers “yes, cyclists are allowed to be here, even in the middle of the lane!”, and also to provide a cue to cyclists to ride away from the door zone (if the sharrows are placed properly).
I’m also surprised that anyone can be surprised that traffic circles are dangerous. That said, the photos on the slides show roundabouts and not traffic circles (is that a difference in dialect?) Roundabouts are actually supposed to be safer because they force drivers to slow down, meaning that even if they crash, the crash is much less likely to be fatal than a crash at an intersection or even at a traffic circle.
ITB, they do show traffic circles. Traffic circles are both the giant Columbus Circle kind but also those very small ones where you barely have to turn. Roundabouts and rotaries are different. Its confusing, and its not clear if the study does take into account the difference.
The giant Columbus Circle kind are unmitigatedly awful.
The micro-mini “one-foot-wide blue hump” kind (which they have in the UK) are also pretty bad.
The small but not miniscule kind have their place.
I like sharrows when they are on side streets, especially as part of a fully thought out “bicycle boulevard” system, as in Berkeley. In New York City sharrows are usually found on slower, lower traffic side streets as well.
Putting them on fast, wide main roads is a mistake.
I’m looking at the study as published in AJPH and it states pretty clearly that shared lanes on Major street route, with parked cars, have a lower odds ratio of crashes than Major street routes with no parked cars. Where there is a difference is on “Local street route,” contrary to what Joseph E. states in the comments above. On those streets, there was a greater odds ratio when there were sharrows (i.e. “Designated bike route with traffic calming”) than on “no bike infrastructure” streets.
The bigger problem with the study is that the participants were drawn from the relatively shallow pool of “those whose injuries were serious enough to result in a visit to a hospital emergency department, but not to cause death or a head injury so severe that the trip could not be recalled.” Cyclists who don’t get into accidents were not included.
What if it is a physically constrained gap in street network between other blocks with bike lanes or cycle tracks? True, “sharrows” are not much more than the striping equivalent of “share the road” signs. But where a network gap isn’t going to see a capital project anytime soon, it helps to sign something for continuity.
Plus, you assume that parking is present on streets with “sharrows.” Maybe the street is just two wide travel lanes. When not having the minimum dimensions for a bike lane (even absent parking), would you then rather have a narrow striped shoulder of fully mixed travel lanes?
While not a comment on safety, I appreciate being able to ride in a new neighborhood and quickly locate the next street in the bike network by seeing the sharrows extending into the distance. Of course there may be other ways to do this (like with bike LANES, colored streetsigns (hidden in foliage)).
Jonathan is clearly right in pointing out the sample’s limitations; by limiting the sample size to bikers who ended up in the emergency room (but not the morgue), we can safely say that the study is bound to overrepresent riskier biking practices, those more likely to send bikers to the emergency room (and yes, the morgue), than the norm.
That said, there are several interesting finds:
1. Odd as it may be, sharrows that mark major bikeways on quieter and/or narrower streets MAY be systemically less safe, and therefore useful, than those marking bikeways in concert with major streets. This is counterintuitive, and undoubtedly merits significant further study, both in terms of better-balanced samples and before-and-after safety studies.
2. The best way to reduce cyclist risk is to provide major bike infrastructure (as has long been the practice in Europe). Cycle tracks and buffered and unbuffered bike lanes are always going to be the better choice, from an engineering perspective, and sharrows should only be implemented when there are significant physical hurdles (and I’m not talking about excessively wide driving lanes here) barring any other treatment.
3. Something interesting is going on with traffic diversion vs. traffic calming. For cyclists, diversion is superior to calming, but this may be because a diverted network effectively creates a pair of mated hierarchies, further forcing auto traffic onto major arterials (which are already known to be, by far, the least safe of road types). A superior understanding of how European networks work may help mitigate this to an extent. Further study here, too, is needed.
4. Traffic circles are significantly more dangerous TO CYCLISTS. Is this because oncoming traffic does not stop but merely yield (an instruction that seems to be unknown to many American drivers)? Perhaps there are ways to make cyclists more visible when dealing with this type of intersection control? Or perhaps traffic diversion is superior along major bikeways, but calming superior away from them?
5. Can there be any doubt anymore that salmoning is the single riskiest biking practice cyclists regularly use?
There is a paucity of good research on bicycle facilities. The ones available have design flaws, small sample sizes, and/or unwarranted assumptions. However, this study does move forward our knowledge. A more appropriate headline would have been something like “small study shows variable safety of different bicycle facilities.” I don’t think “Sharrows Suck” is warranted by the research study.
Sharrows that are placed to the side next to parked cars are worse than useless. They encourage dangerous behavior. Unfortunately, that’s the type of sharrow used in the study. The study set up sharrows to fail.
Sharrows need to be placed properly in order to be effective. They need to encourage bicyclists to control the lane — not try to share it side by side with cars as these did an which are dangerous. They should be centered between the tire tracks of where the cars typically go in the lane.
The people who say that sharrows don’t work are the same people who have never really put in the effort to properly study vehicular cycling, much less do the mileage using the techniques that you need to truly understand how it works.
Seriously agree. I can’t stand sharrows, I wrote a post about them a while ago too… I don’t see how they’re any better than nothing at all. All roads can be used by bikes, so they’re really just there to remind drivers. But sometimes I think that they tell uneducated drivers that only roads that are marked are for cyclists.
[…] Engineer’s” transportation satire blog, Systemic Failure, for some insight: Sharrows Suck and Sharrows Are Not A Bike […]
[…] Engineer @ Systemic Failure minces words with: Sharrows Suck . I wish he would tell us how he really […]
[…] https://systemicfailure.wordpress.com/2012/11/14/sharrows-suck/ […]
[…] will be prioritized for the space that remains afterwards. Cycle-users will be lucky if even “Sharrows” remain in areas around these […]