Top: Home Page
Up: Cambridge bicycle facilities and program

John S. Allen
7 University Park
Waltham, MA 02154-1523
(617) 891-9307 voice/fax
Technical writing, translation
Mechanical design, acoustics
Consultant on bicycling
Effective Cycling instructor

Comments on
Cambridge bicycle policy draft statement
by John Allen

Originally submitted July 1995; revised version of January 17, 1996

In general: I think this document makes a good and positive statement, but there are many technical details that need refinement—so read on! You may assume that I agree with and support sections on which I do not comment.

p. 5 Statement about reduced number of costly accidents and property damage is only partly true, since the risk of injuries and fatalities per mile of travel is greater for bicycling than for public transportation or private motor vehicles (except motorcycles). What can be stated truthfully and positively is that the health benefits of bicycling greatly outweigh the accident risk, that bicycling is safer than inline skating or walking per mile of travel, and that the risk per hour of exposure is about the same for bicycling, motoring or walking.

The words “modifying the existing street system to rededicate some of the space to nonmotorized modes” open the entire Pandora’s box of bike lane design issues. In most cases, shared us of road space is more practical, with wide outside lanes open to both bicyclists and motorists.

The statement that “a person in reasonable physical condition can ride a bicycle up to at least three miles -- five miles is considered reasonable bicycling distance -- with minimal physical exertion” is confusing. Generally, bicycling is as strenuous as walking at 1/3 the speed. At 10 mph, bicycling is no more strenuous than walking. Anyone except a very old or infirm person who rides a bicycle regularly enough to get minimal training can maintain 15 mph for hours. The limit on bicycling distances is not physical capacity but available time.

p. 6 “people have repeatedly stated their desire for more and better bikeways (and walkways.” Most people don’t know what a better bikeway is. Following their wishes has led to the great majority of bikeways’ being poorly designed and located.

Rodale study was conducted in 1990, not 1992.

p. 8 again shows a strong bike lane bias without spelling out what is an acceptable bike lane. The bias here is to promote bicycle use, not necessarily to build good facilities.

p. 9 The Worldwatch pamphlet by Marcia Lowe has been devastatingly analysed by John Forester. He maintains that most of what it says is untrue even if it sounds good.

Here the Rodale study is cited as being from 1991. This is probably the publication date rather than the time research was conducted. This study has been republished by the US CPSC in its “Bicycle Use and Hazard Patterns in the United States (1995).

p. 10 and 11 The definitions here are not those under Massachusetts law and regulations. See Chapter 90 definitions, 720 CMR definitions and 350 CMR (MDC) definitions.

In particular:

The definition of “bicycle” here is awkward and much less good than the one under Mass. law, which stresses the pedal power aspect, is inclusive of tricycles including adult tricycles and avoids the confusing wheel size issue.

The definition of “bikeway” is misleading, because there are very few “bikeways” from which pedestrians are excluded, as a practical matter. The only exception that works at all is when a pedestrian path parallels a bicycle path as in Danehy Park.

Definition of MUTCD refers to “State of Massachusetts.” Massachusetts is a Commonwealth, not a state.

The definition of “vehicle” is quite unlike the one in 720 CMR or 350 CMR.

Also, Massachusetts uses some unique terminology about roadway and right of way.

p. 13 Under the heading “Motor Vehicle-bicycle crash summary”, we read that approximately 75% of bicycle crashes do not involve motor vehicles. What is the source for this? Every study I’ve seen puts the percentage between 80 and 90%.

Then the section on accidents goes on to discuss almost nothing but bicycle-motor vehicle collisions. Shouldn’t we be as concerned about the other 80 to 90 percent of crashes? Don’t they matter?

Colorado Dept. of Transportation draft design guidelines: do the conclusions here reflect original research? They seem to be drawn from the classic Cross and Fisher study from around 1980. Note that the CTPS conducted a study about 10 years ago in the Boston area which closely confirmed Cross and Fisher. Cathy Lewis has details. This document should cite it.

The Colorado statistics do agree with Cross and Fisher about the two accident types which are made more likely by bike lanes.

University of North Carolina: this analysis appears to reflect original research. However, some of the statistics lack context at least as presented here. The accident rate on different types of roads, or with on-street parking, can not be determined without knowing the amount of riding there. When referring to alcohol use, is this by the cyclist or by a motorist who struck the cyclist or a pedestrian who staggered out in front of the cyclist? Another recent study from Johns Hopkins University shows that a high percentage of fatal cycling accidents involved drunk cyclists!

p. 14 Oregon State Bicycle Plan: this detailed analysis treats only the 10-20% of bicycle crashes that involve a car. Also note (p. 15) the contradiction between the first and second paragraph as to who is at fault, and in the third, the term “invisible to cars.” Bicyclists are never visible to cars since cars can’t see. Generally, their drivers can, or else they would run into something before they even got out of the driveway. The confusion of cars with their drivers is a very common one, typical of people who do not think clearly as bicyclists about how drivers interact with them. It is a strong symptom of what John Forester calls the “cyclist inferiority phobia.”

p. 16 The conclusion that bikeways increase safety and traffic law compliance in Eugene and Corvallis, Oregon is not justified here. Simply juxtaposing the facts that accident rates are lower and that there are bikeways does not lead to this conclusion. Couldn’t education and law enforcement in these bike-conscious towns have some effect? France has very few separate bicycle facilities, yet all bicyclists obey the rules of the road. That is because they are taught in elementary school and they are enforced. The careful Wachtel and Lewiston study in the September, 1994 ITE Journal shows no measureable effect of bike lanes on safety.

p. 17 should say: “Providing shoulders of equal width and quality on each side of the road.”

p. 18: the section on “Hazards of routing cyclists off thoroughfares onto less-traveled streets” is easily read as a justification not to improve cycling conditions on secondary streets. This is where most people live and most children ride. I treat this topic in more detail in my statement of January 17, 1995 to the Committee.

p. 19 Definition under “Brakes” doesn’t reflect Massachusetts law, and is not technically correct, since the front brake can lock but not skid the wheel on dry pavement.

Definition under “Lights” seems to say that white reflectors are used on the rear. A red reflector or taillamp is what the law requires.

p. 20 Now once again we say that most bicycle crashes do not involve motor vehicles. But there has been no discussion in this section of how facilities can help prevent other types of crashes e.g. bike-bike, bike-ped and single-bike.

Research (Harborview Medical Center, Seattle) shows that helmets reduce head injuries by 88% and brain injuries by 85%, not “up to 80%.”

All in all: this section is built on a weak foundation, does not discuss 80 to 90% of accidents, and could do well by referring to some of the classic studies such as Cross and Fisher, Kaplan, Chlapecka, Shupack and Driessen—Forester summarizes some of them in his section on bicycle accidents in Effective Cycling (MIT Press).

p. 21 I’m not comfortable with the idea that bike lanes “are to be provided wherever possible.” for several reasons.

1)    Double-parking conflicts and bicyclists’ being made a scapegoat for increased enforcement against double parking. This is a great way to make enemies for a bicycle program.

2)    The idea that if there is a bike lane, that is the only place bicyclists should ride. Especially troublesome in combination with double-parking in the bike lane...

3)    Intersection conflicts and misunderstandings, most commonly bicyclist left turn from right lane and motorist right turn from center lane. Bike lane stripes must be dashed or dropped before intersections.

All in all, I’d prefer undesignated guide stripes or wide outside lanes in most places. Bike lanes can make sense in some, such as on bridges or in other long stretches without cross streets or driveways.

p. 23 refers to “Massachusetts State Law.” In fact, bicycles are defined as vehicles in the Commonwealth of Massachusetts Regulations 350 and 720. Also, where bicycles are prohibited is explicitly spelled out in the regulations and also in the Statutes (Chapter 85, section 11B) as only on limited-access or express state highways, which must by definition be divided highways, generally with grade separations at intersections, and where signs are posted. In other words, bicycles may not be prohibited anywhere in Cambridge except maybe on parts of Memorial Drive, the McClellan Highway, and Route 2. In practice, bicycle prohibitions are not posted there, either.

“A wide outside lane that accommodates both cyclists and motor vehicles is one way of designating a shared roadway.” It is also often one way of improving a shared roadway either by widening it or by restriping. “These lanes should be provided where there is inadequate width to provide bike lanes.” Once again, this document’s astonishing pro bike lane bias surfaces. Many, and especially cyclists, would argue that a wide outside lane (perhaps with guide lines but not specially designated for bicycles) is preferable to a bike lane under most conditions for a number of reasons, in particular because a wide outside lane provides the same amount of additional travel space without creating confusion about the traffic rules: read on.

p. 24 “The appropriate facilities for bicyclists are bike lanes on urban arterial and collector streets.” This is a very sweeping and one-sided statement with which many cyclists strongly disagree and which is open to serious scientific questioning. Wachtel and Lewiston (ITE Journal, Sept. 1994) show no difference in accident rates between bike lane and non-bike lane arterials. There’s also the question of placing all your political eggs in this one basket, which has resulted in political failure e.g. in New York. The issue of bike lanes has been thoroughly studied and debated in California, where they have been in place for 20 years. Let’s benefit from what has been learned there. Consider the following comments a counterbalance to the heavy pro-bike lane bias of the draft document. I am not opposed to bike lanes where appropriate, but let’s be reasonable about them, not discuss them as if they were paved with gold.

Most of the statements in support of bike lanes in this section are false or only partially true:

“Bike lanes are appropriate on minor collector or local streets if traffic speeds and volumes are higher than...25 MPH or ADT over 3000.” The justification for bike lanes depends also on the volume of bicycle traffic. A designated bike lane is required to be 4 to 6 feet wide because the engineering assumption is that the bike lane must be wide enough for 2 bikes, one passing another. That way, bikes going straight through never have to use any other lane. That approach makes sense when motor traffic is very heavy and fast and bicycle traffic is also heavy, but, considering the need to add 4 to 12 feet of additional pavement or to remove it from other uses, is very costly under less demanding conditions. Bicyclists can find opportunities to overtake without having a lane all their own. On most Cambridge local and collector streets, bike lanes would require removing either a travel lane or a parking lane. This an expedient which should be used with discretion, where unavoidable, to avoid political repercussions.

“Bike lanes help define the road space: true.” But the definition provided is controversial.

“...provide bicyclists with a path free of obstructions”: untrue. This is accomplished by strict law enforcement against double parking, not by lane striping. Parking in bike lanes is an endemic problem where enforcement is less than strict. Illegal parking is an endemic reality and serves the unavoidable need for businesses to receive deliveries. Eliminating parking in bike lanes in Cambridge would create such a reaction from the business community as you would not believe.

“...decrease the stress level of bicyclists riding in traffic”: true, as long as the bike lanes can be kept clear of obstructions, and as long as the bicyclist is not attempting to overtake or turn left in a travel lane narrowed by a bike lane on its right.

“...Encourage bicyclists to ride in the correct designated roadway position”: True if they are going slower than other traffic and not turning left, and going the right way. Bike lanes have been shown to encourage wrong-way riding, however.

“...and signal motorists that cyclists have a right to the road”: This is simply untrue. Bike lanes signal motorists that bicyclists have a right to the bike lane, and no right to the rest of the road. This becomes particularly onerous when, as often follows, laws or ordinances are passed that eliminate or decrease bicyclists’ right to the rest of the road. The first question an insurance company lawyer asks when a bicyclist was injured on a bike lane street is “was the bicyclist in the bike lane.” If not, the case is very rarely winnable before a jury even if the bicyclist was legally in the right.

“Bike lanes are intended to promote an orderly flow of traffic, by establishing lines of demarcation...” This is a waffling statement: it does not say that bike lanes actually do promote an orderly flow of traffic. The standard traffic principles are channelization by speed between intersections, and channelization by destination at intersections. In fact, the bike lane system conflicts with the ordinary rules of the road under the following conditions: motorist turning right; bicyclist turning left; bicyclist overtaking motorist.

“Bike lane stripes can increase bicyclists’ confidence...” So can guide stripes without a designated bike lane, but no claim for this was made in the earlier section. Let’s be consistent. This is an argument for striping, not for 4’ to 6’ bike lanes.

“Likewise...passing motorists are less likely to swerve toward opposing traffic...” This statement should not be made without a research citation. In practice, the need of motorists to swerve toward opposing traffic depends fundamentally on the motorist’s expectation of cyclist behavior and the width of the road. Cyclist behavior varies. Where cyclists ride predictably, motorists don’t give them unnecessary room, bike lane or no bike lane. This is more a problem of education than of engineering, in my opinion—reinforced by riding in France, where cyclists ride predictably. In any case, motorists are required to drive closer to opposing traffic by a bike lane even when no bicycles are present. This inflexibility, taking discretion out of the driver’s hands and making it law, also can lead to crashes.

“NB: The presence of a bicycle lane does not preclude the bicyclists from riding in the motor vehicle travel lane if the situation warrants...” . Even if this rule is codified in the law and understood and observed in the community, most bike lanes on Cambridge streets would have to be constructed by narrowing or removing travel lanes, not by adding pavement width. Therefore the maneuvers which must be made using the travel lane—overtaking and preparing a left turn—may become more difficult and dangerous. As a practical matter, a redneck attitude toward bicyclists is very common in the US, and many motorists will be angry to see bicyclists anywhere except in the bike lane. Besides, this misstates the law. There is no such thing as a motor vehicle lane. The lanes other than the bike lane are as open to bicyclists as to motorists under Massachusetts law.

“they establish the correct position of bicyclists on the roadway” “Establish” ? The law, not lane lines, establishes this particularly in the light of what I’ve just said. Delineate, perhaps, and sometimes a correct position (there is no single correct position), sometimes an incorrect one.

“They promote an orderly flow of traffic” They may increase order in some ways but they increase disorder in others, as discussed above.

“they allow bicyclist to pass motor vehicles stopped at a signal (coming up to the stop line, and also stopping for the signal).” Motorist right turn in front of bicyclist is one of the top 2 or 3 causes of bicycle accidents, as cited on page 13 of this same document and mentioned again on p. 26.. Bicyclists should never come up to the stop line at an intersection where right turns are permitted. Facilities certainly should not encourage this.

“they send a message to motorists that bicyclists have a right to the roadway.” I’m sorry to have to say so, but this is sophistry. As discussed above, they send a message that bicyclists have a right to the bike lane.

“they remind motorists to look for bicyclists on the road” Once again, we have a solution looking for a problem. It could as easily be stated that bike lanes let motorists be heedless of bicyclists anywhere except in the bike lane, because motorists assume that all bicyclists will be in the bike lane.

“they give bicyclists a clear place to be so they are not tempted to ride on the sidewalk” True, but that doesn’t mean that bicyclists can simply ride in the bike lane and not have to learn correct maneuvering according to the rules of the road. Bike lanes, as discussed above, can encourage bicyclists to make incorrect maneuvers. There is no solution that eliminates cyclists’ need to be knowledgeable and to take responsibility for their own safety.

reverse of p. 26: note that intersection diagram shows motorist merging into the bike lane to make right turn, contrary to what Cara Seiderman has advocated.

p. 26 “a two-legged right turn...” fails to state that this can be made in safety only by coming to a full stop at the far corner of the intersection.

“It is also desirable that detectors in left-turn lanes be sensitive enough to detect bicycles...At intersections without bike lanes, it is desirable to install detectors that are sensitive enough to detect bicycles.” The acknowledgment that detectors should detect bicycles is welcome. But it is not only desirable. It is mandated by the law which defines bicycles as vehicles.

p. 27 Treatment of bike paths is too dismissive. There are several existing bike paths in Cambridge and others proposed e.g. Watertown Branch, Alewife Brook Parkway, Charles River Park extension, railbed between Vassar Street and Albany Street.. Existing paths need improvements, and these should be discussed, as well as details of design for proposed ones.

Treatment of traffic calming should be much more detailed, with examples. In particular, this report should address the issue of replacing conflicting one-way signs with other traffic calming measures which are more bicycle-friendly. This is a key element to improving bicycle friendliness of Cambridge’s residential neighborhoods.

p. 29 Right turn only lanes: The same principles that apply to right turn only lanes also apply anywhere motorists may turn right -- any intersection or driveway. This is the Achilles’ Heel of the bike lane approach. To warn against the right turn problem only where there is a designated right turn lane is inconsistent.

p. 30 50 foot dashed length before intersections is far too short to allow a proper merge.

p. 32 Loop and other detectors not only “should be tuned and retuned to detect bicycles.” They should be designed to detect bicycles, using quadrupole or California D-type loops.

Bike path-road crossings are one important location for bicycle specific signals which is not mentioned.

“Assuring the sensitivity of existing loop detectors to detect cyclists.” These are not just “improvements that can be made” but are mandated by the definition of bicycles as vehicles.

Inset before p. 34: Railing shown is of poor design for bicycles, with vertical bars rather than horizontal rub strip, opaque through curve and not set back from inside of curve (bicycles lean to inside of curve).

Appendix F List of statutes cited is incomplete. See the index to the statutes for a complete list. I have a reasonably complete collection of statutes in a notebook. Also, regulations, particularly sections 350 and 720, bear directly on bicycling. Section should also include Cambridge ordinances relative to bicycling, not just parking ordinance.

References This list of references is heavy on speculative studies and light on basic research works. Many important references in the field of bicycle traffic engineering and accident studies are missing. I have mentioned some in my comments. More balance in references would reflect a more balanced presentation is several areas, notably accidents and facility types.

Top: Home Page
Up: Cambridge bicycle facilities and program