One of the pleasures of driving in the United States is that lanes are striped to tell people where to drive—and where to expect that other people will be driving. Most of the time, one can simply keep driving in a lane with little concern that the lane will suddenly become a hazard. To be sure, one must always keep a lookout. But changing lanes is usually optional, something one does before turning, to pass slower traffic, or allow faster traffic to pass. Where there is a potential conflict with other traffic, a traffic light, yield sign, or stop sign tells everyone who has the right of way; and other drivers generally follow the rules. A given sign or striping on the pavement always means the same thing, and the meaning is clear.
One of the most maddening aspects of riding a bicycle, by contrast, is that signs and pavement markings have unclear meanings or mean different things in different situations. Along many suburban roads, it is very difficult for a cyclist to figure out where the transportation department wants her to ride. The drivers let her know that they want her over on the shoulder. But to ride on the shoulder, she must continually cross solid white lines as the shoulder is overlaid with bypass lanes, right turn lanes, and even through-lanes that swerve right. The shoulder may be eliminated at narrow bridges and some intersections.
Someone riding a bicyle must often change lanes simply to continue straight. Read literally, the lane striping generally indicates that the cyclist is supposed to shift right into a ditch, or ride straight into the bridge structure. While the cyclist can assume that this was not the intention, it is usually unclear whether the highway department has not thought about what the cyclist should do, or assumes that the cyclist should stop and wait until there is no traffic, squeeze along the extreme right of the roadway in terror, or ride in the center of the lane (again in terror as fast-moving cars close in upon her). It often seems as if the road agency's perspective is that there may be no law against riding a bike on the road, but bikes are not recognized road users either whose path should be considered in road design or communicated to road users.
What Washington, DC and Most Major Cities Are Doing
Bike lanes can provide cyclists with a clear course to navigate a roadway, with signs and pavement markings that tell both drivers and cyclists what the cyclist must do. Unfortunately, bike lanes have not achieved their potential, because in most jurisdictions the lane markings can mean two different things:
- Along roads where parking is prohibited, the bike lane is a lane in which cyclists are encouraged to ride
- Along roads where cars are parked along the curb, the bike lane is a door-hazard zone where people should generally not ride a bicycle except at very slow speeds. Instead one should ride along the left stripe, which is (usually) slightly beyond the door hazard area.
The purpose of the bike lane may help to determine the appropriate design. As a general rule, a mid-sized SUV’s door extends about 11 feet from the curb. (E.g. the SUV is 6-1/2 feet wide, the door opens 3-1/2 feet, and is parked 1 foot from the cub). If a bicycle is 2 feet wide, then if its tire is 12 feet from the curb, the bicycle is riding along the outer edge of the door zone. As it happens, the most widely used standard for bikes lanes recommends that bike lanes extend at least 12 feet from the curb. (See American Association of State Highway and Transportation Officials. 1999, Guide for the Development of Bicycle Facilities.) Thus, a cyclist who wants to be safe from opening car doors might ride just to the left of the leftmost line of the bike lane. Although people call the area between the two white stripes a “bike lane”,” in the case of the AASHTO standards it would be more accurate to say:
- The right stripe is the “parking line”
- The left stripe is the “bike line”
- The area between the parking line and the bike line is the “door hazard zone.”
Viewed in this context, it is apparent that the bike lane along curbside parking is very different than the bike lane along an open roadway. Rather than painting a bicycle symbol, these bike lanes would more accurately have diagonal striping to show that vehicles ought not drive there. Of course, that is not going to happen.
Providing cyclists with a safe lane for travel was not necessarily the main goal of bike lanes. Many advocates and municipal transportation departments had a more modest goal: providing a safer lane for travel. Many cyclists ride in the door zone anyway, so creating a lane to delimit this zone may not increase the hazard for those cyclists. For those who are trying to stay out of the door zone, the bike line is quite useful. And if many less informed cyclists feel safer having a lane, the bike lanes might increase the amount of cyclists on the road, which might make the average cyclist safer—though perhaps not the new cyclists lured onto the road by the illusion of a safe bike lane. Finally, bike lanes may help to keep cyclists out of the way of motorists, or at least provide motorists with the illusion that they are doing so.
Efforts to Mitigate the Hazard
The evolution of standards for bike lanes shows a trend toward increasing recognition of the door hazard problem. Even the 1999 AASHTO standards recognized the hazard to some extent, suggesting that the bike lane be farther from the curb where there is a high turnover of parking. The risk-benefit logic was that that the risk of getting doored or struck by a vehicle pulling into the lane is acceptable along a parking space where a car is boarded 1-2 times a day, but possibly not acceptable along a space where a car is boarded once an hour.
The logic for worrying more about door zones with more opening doors is clear, but no analysis was ever conducted on where to draw the line: If each car door randomly opens once every 8 hours, then for every 8 hours of riding in a door zone bike lane, one door will open when you are less than 1 car length away, another door will open when you are 1-2 car lengths away, another door will open when you are 2-3 car lengths away, etc. If 10% of people fail to look before opening the door, then those are the risks for every 80 hours of riding in an AASHTO bike lane. Is that an acceptable risk for every 80 hours? Or should it be 800 hours or 8000 hours? They did not analyze that question, though the logical inference from a standard that does not change the status quo is that the status quo is acceptable: in effect, AASHTO implicitly decided that the existing rate of door hazard injuries was acceptable; traffic engineers ought not create worse hazards than the norm; but it is not worthwhile to decrease the hazard either.
By 2006, Maryland’s State Highway Administration added an extra 1-2 feet, requiring the bike lane to be at least 5 feet and the parking lane 8 feet, with a recommended width of 9 feet, which would leave the left edge of the bike lane 14 feet from the curb. In 2010, the National Association of City Transportation Officials recommended that bike lanes extend 14.5 feet from the curb, and that the bike lane be at least 5 feet wide, but preferably 6 feet. With a NACTO bike lane, one would be beyond the SUV’s door zone as long as one is riding in the left half of the bike lane, which NACTO illustrates with this photograph. A draft revision of the AASHTO standards suggested that lanes as wide as 7 feet would be appropriate where there is a high turnover of parking (e.g. restaurants).
Meanwhile, the education of cyclists often emphasizes the need to avoid riding in the door zones. Unfortunately, those efforts do not reach all cyclists. The existence of half the cyclists riding in door zone bike lanes while the other half ride just outside the bike lane reinforces the impression of drivers that cyclists are unpredictable, arbitrary scofflaws. Consider the AAA towtruck driver, who struck a cyclist and then told the police with surprised outrage that the cyclists weren't staying in the bike lane. The dichotomy of cyclist behavior parallels what we see along narrow roads without bike lanes, where some cyclists hug the curb while others take the lane. Except in that case, most drivers understand (if reluctantly) that the cyclist is using the full lane because the road is narrow. Why would it ever occur to the average driver that the bike lane is where bikes are not supposed to ride?
What Should Prince George's County Do?
Cycling experts almost universally agree that cyclists should never ride at full speed in a door zone. Therefore, Prince George's County should never build a bike lane that encourages someone to ride in a door zone. If the only choice is between a bike lane in the door zone and no bike lane, then the County should not stripe the bike lane. Instead, the County should paint sharrows and install signs that say “Bicycle may use full lane.”
Because the door zone of an SUV extends 12 feet from the curb, a roadway would need 16 feet for a 4-foot lane entirely outside of the door zone—5 feet wider than needed for the typical 11 foot lane used for parking (or right turn lanes near intersections) on county roads. Road agencies are usually reluctant to make the roadway 10 feet (5 feet in both directions) wider just for bikes. (Most roads with bikelanes without parking would have had a shoulder or wider lane were there not a bike lane.) But there are some options for creating bike lanes when less space is available, most of which fall into two categories: Marking the door zone, and limiting its reach.
Conceptually, the problem with the current approach to striping bike lanes is that the right stripe serves two incompatible functions: the right edge of the bike lane and the left edge of where a car tire can be parked. The door hazard zone is a third zone whose location must be guessed unless it is marked. Maryland SHA’s design manual refers to some pavement markings for showing the door zone within a bike lane. Several cities have tried different approaches to either define the door zone within the bike lane, or define the door zone as not part of the bike lane.
Alternatively, the reach of the door zone can be limited by pushing the cars closer to the curb and limiting the size of vehicles that park next to a bike lane. Design guidelines for bike lanes have sometimes minimized the door hazard by depicting unusually small vehicles in their design drawings. The most famous case was in Chicago, where design drawings used a pickup truck that was only five feet wide. The SHA manual assumes that the door only reaches 9-1/2 feet from the curb instead of the 11 feet we assume for an SUV; the entire vehicle is within 7 feet of the curb and the door only extends 2-1/2 feet. Although this situation is optimistic for unrestricted parking, if the parking lane is striped at 6-1/2 feet and/or signs are posted restricting parking to compact automobiles, it may be realistic to limit the door reach to 9-1/2 feet from the curb. If a bike lane was then striped at 12 feet from the curb and the door zone marked at 11 feet (instead of 9-1/2 feet to reflect the width of a mountain bike and provide a few inches of error), then the bike lane could delineate a one–foot wide lane in which riding would be outside of the door zone. If 14 feet is available, that lane could have a three foot travel lane entirely outside of the door zone.
Thus, a wide variety of options are available to create bike lanes that discourage riding in the door zone, even when space on the roadway is limited. If those options are not practical, then the County should not stripe a bike lane until the road is widened or the main travel lanes are narrowed.
If bike lanes in door zones are good enough for the District of Columbia and other jurisdictions, shouldn’t they be good enough in Prince George's County? Not necessarily. In Maryland, § 21-1205.1 of the Transportation Code requires cyclists to ride in the bike lane, which creates a reciprocal duty that bike lanes be designed to be safe places to ride. (Cyclists are allowed to leave the bike lane to avoid hazards, but the law clearly evinces a legislative intent that the lane itself not be the hazard.) Moreover, county roads often have stretches with parking and stretches without parking: whatever the merits may be of striped bike lanes meaning two different things in a given jurisdiction, it would be a really bad idea for a given traffic control device to mean two different things on the same road, which would be the case if the bike lane were a safe place to ride in some places and a door zone elsewhere. Furthermore, the density of cyclists is much greater and the traffic speeds much lower in the District. Bikes riding outside of door-zone bike lanes in the County would tend to provoke rage from uninformed drivers, while in the District it is more likely to seem like a tolerable, mysterious convention as long as the cyclist is riding parallel tio the bike lane.
Finally, we now have the option of painting sharrows where there is insufficient space for a safe bike lane. In the past, officials may have felt that a door-zone bike lane would still be better than no pavement markings that directed and recognized the legitimate rights of cyclists to the road.
Going forward, bike lanes should be designed to actually be what drivers and most cyclists think they are—the best place to ride a bicycle on a given street.
(Jim Titus is on the Board of Directors of WABA, Maryland's Bicycle and Pedestrian Advisory Committee (MBPAC), and the Prince George's County Bicycle and Trail Advisory Committee (BTAG). The opinions expressed herein do not necessarily represent the official opinion of WABA, MBPAC, or BTAG.)