Bicycle Facility Selection: A Comparison of Approaches
by Michael King, Architect
Bikelane Design Guide by Chicago DOT
Published by: Pedestrian & Bicycle Information Center, Highway Safety Research Center,
University of North Carolina-Chapel Hill

9 November 2002

home                    facilities

1 Authorities

These two documents are published by the authority of Charles Zegeer, Director of the Pedestrian & Bicycle Information Center, Associate Director of the Highway Safety Research Center, University of North Carolina, and of Andy Clarke, Technical Advisor to the Pedestrian and Bicycle Information Center, and Executive Director of the Association of Pedestrian and Bicycle Professionals. In short, these two documents ought to represent the best that the bike planners can do, so that they should be evaluated accordingly.

2 Bicycle Facility Selection

2.1 Quality of Information

One would presume that these two documents would be roughly compatible, coming from the same source at the same time. However, they are not. King states: "Tucson AZ has bike lanes on 40-foot wide residential streets with volumes around 200 ADT. 40-foot wide streets in Chicago have four lanes and carry up to 40,000 ADT, yet have the same size bike lanes." The statement raises two issues, accuracy of claim, and possibility of accomplishment. The 40-foot Tucson bikelane streets presumably provide either 7' parking, 4' bike lane, 9' motor lane each side, or 8' parking one side, with 12' motor lane, 4' bike lane each side. The Chicago Bike Lane Design Guide lists no streets of 40-foot width with bikelanes.

 The narrowest streets with bikelanes listed in the Chicago Bikelane Design Guide are 44-foot width, with two motor lanes and parking, the layout being 7' parking, 5' bike, 10' motor on each side. The Chicago Guide states: "The City of Chicago does not stripe a bike lane of less than five feet in width." Now consider the possibility of King's claim. In 40 feet he gets 4 motor lanes and 2 bike lanes. How would one do that? The layout would have to be 8' motor lane, 8' motor lane, 4' bike lane each side. Considering that lawful vehicles are up to 8'-4" (100") in width, 8-foot motor lanes are just not practical, and, so far as I know, no streets have ever been so striped. If the bike lanes were striped to what Chicago states to be its minimum of 5', the motor lanes would have to be only 7.5' each. In short, Mr. King has no more idea of traffic engineering and its bicycling aspects than one might expect of an architect.

2.2 Purpose of Document

King has surveyed 16 bike-lane design guides that discuss bikeway design in terms of the speed and volume of motor traffic on the adjacent roadway. It might have been possible to grant King's research some credit, as long as it is considered to be only a summary of the errors of bike-way planning around the world. However, King assumes that these guides possess a sufficient degree of scientific credibility, while never discussing that issue. "Ultimately a goal of this exercise is to produce a design tool that may be used by the bicycle and pedestrian design community to obtain higher quality walking and cycling facilities - minimum standards. [emphasis in original] ... Just as other professions have minimum standards, why should not the pedestrian/bicycle community not produce minimum standards for their facilities?"

 Mr. King does not understand that bicycle planning is not a profession. Bicycle planning is the business of conning money out of governments by using false claims of engineering knowledge about bicycle transportation. However, the knowledge that does exist about bicycle transportation directly contradicts the claims of the bike planners. Mr. King has not troubled himself to examine whether or not there is a scientific foundation for the data that he has spent so much trouble over. Neither has his publisher, the Highway Safety Research Center of the University of North Carolina. Of the few of King's sources that are familiar to me, I know that there is no scientific foundation for them, and I have never heard of any scientific foundation for any such schemes. In other words, Mr. King's work is merely another storey in bike planning's house of cards. That conclusion reached, why should anyone bother to read any more? Only because the political power of the bike planners necessitates that not only the corpse be shown to be dead, but the appropriate stake must be driven through its heart to prevent its spirit from arising from the grave.

2.3 Methodology

King collected 16 examples of documents that related width and type of bicycle facility with volume and speed on the adjacent roadway. He classified the bicycle facilities into: 1) Narrow traffic lanes, assuming 11 foot width. 2) Wide traffic lanes, assuming 14 foot width. 3) Bike lane, assuming 5 foot bike lane alongside 11 foot motor lane. 4) Separated lane, either wider than 6 feet, or with some kind of separation. "It may mean a wide bike lane with median-type striping ala New York City. It may mean flexible bollards like those found in Montreal. It may mean a Danish `cycle track' - a bike lane raised halfway up the sidewalk curb. It may mean simply an eight-foot wide bike lane, as found in Davis CA." King then grouped the speed and volume of motor traffic criteria of each document into a unified group system. Then, King listed the groups of motor traffic against the groups of bikeway types. Naturally, there is some degree of agreement among nations: more intense motor traffic, more definite separation. King combines the data from all nations for each type of facility into his estimate for the typical relationship for that type of facility. King then plots his motor-traffic groups against the American Bicycle Level of Service. BLOS is a measure of the degree of fear engendered in incompetent cyclists by the presence of same-direction motor traffic. This is purely a subjective measure of superstition and inexperience. King is then able to draw, on the plot of all-nations' data for each type of bikeway, a set of boundaries for the different levels of BLOS, and therefore to estimate the BLOS values provided by the different nations.

2.4 King's Conclusions


From this system, King draws conclusions. "From this we see that North Americans generally provide LOS D while the Europeans and Australians provide LOS C. Regardless of location, narrow and wide lanes generally provide LOS D, bike lanes LOS C and separated lanes LOS B."

 In other words, Europeans provide greater separation, generally in the form of the sidepath that the USA does not specify. Nothing new in that. The other conclusionary statements that King makes within his paper are equally demonstrative of ignorance. His grouping of wide bike lanes with sidepaths as one type of facility shows how little he understands about traffic, for the two types require radically different operating procedures. His statement about integration of bicycle traffic is as topsy-turvey as one might expect from one so imbued with the cyclist-inferiority superstition. "Guidelines from the US and Canada are more planning oriented - about how to increase cycling and provide more bike-specific facilities. Overseas guidelines are much more inclusive in terms of seeing the bicycle as an integral part of the transportation system - making each street safe for cyclists." In other words, kicking cyclists off the roadway onto sidepaths is seen as integrating cyclists into the transportation system!

2.5 Reasonable Conclusions

It is reasonable to examine the claim that the speed and volume of same-direction motor traffic is determinative regarding bicycle transportation. Car-bike collision statistics show that same-direction motor traffic intensity has very little connection with car-bike collisions. Traffic-movement analysis shows that it has very little connection with the skill of traffic cycling, because all that the cyclist does is to ride straight whatever the intensity. Consider Mr. King's use of same-direction motor traffic intensity. He specifically assumes two-lane roadways with 12,000 maximum ADT. That volume is the approximate maximum for such a street. But his tables have 41 entries of volumes exceeding 12,000, going up to 40,000. Such volumes require multiple-lane roadways. Even if we assume some truth to the fear of same-direction motor traffic, there is no reason to fear the motor traffic that is operating two or three lanes away from the cyclist. Mr. King should have recognized this in his analysis. Consider Mr.King's evaluation of the BLOS of separated bike facilities. Just as with bike lanes, he gives these separated bike facilities lower BLOS ratings according to the intensity of the same-direction motor traffic. I see no reason to do this. Paradoxically, if one chooses to believe the superstition, the higher the intensity of the motor traffic avoided, the higher should be the degree of service rendered by the separated facility. In fact, it probably is possible to provide a BLOS for a separated facility, in accordance with the number of roadway intersections, frequencies of children and dogs, proportion of incompetent cyclists, and the like, but these are entirely distinct from the intensity of same-direction motor traffic on the adjacent roadway. Consider the realism of the BLOS system itself. Both I and my partner cycled to work for years, often separately, on a 40-foot wide road that was 2-wide-lane with parking in some places and 4-lane-narrow-lane without parking in others, carrying between 10,000 and 15,000 motor vehicles per day at nominal 25 mph. The BLOS rating for such conditions is E. We had no trouble at all, just delightful cycling along the route that we preferred unless we wanted more hills on the homeward trip. Mr. King's paper is just another piece of bike-planning propaganda, based on nothing more than superstition but tricked up with all the trappings that make it appear to be a piece of valid research. One more problem for vehicular cyclists to counter.

3 Chicago's Bike Lane Design Guide

3.1 General

This document provides standard plans for 18 intersection sizes and 6 bus stop situations. It also specifies striping, marking, and signing for bikeways. Only one detail is new to me. That is, the standard design for lane tapers where the bike lane either disappears into the "motor" lane, or the bike lane starts from a "motor" lane. Chicago provides a parking stripe as well as a bike-lane stripe, in order to keep parked cars closer to the curb. Also, Chicago uses parking lanes of 7-foot and 7.5-foot width as well as the standard 8-foot width, depending on the roadway width available.

3.2 Some Details

3.2.1 Pictorial Misrepresentation

The standard plans carry dimensioned lane widths. They also depict cars and bicycles using the lanes. The pictorial appearance is that there is plenty of room. However, that is misleading, because the drawings have been incorrectly done. It is easy to calculate the scale of such a drawing. Say the drawing gives a lane width of 10 feet, and the width measured on the drawing is 23.5 mm. The scale factor is 2.35 mm/foot. If the drawing of a car is 11.5 mm wide, dividing 11.5mm by 2.35 mm/foot gives a car width of 4.9 feet. These are the actual values for the 44-foot roadway drawing and the one car shown moving in the motor lane.

 The widths of the cars shown in these drawings range from 4.4 feet to 5.4 feet, with a typical width of 4.9 feet. Excluding the smallest cars such as the Geo Metro, the width of cars ranges from 5.8 feet to 6.1 feet, with vans running 6.6 feet. In other words, the cars shown in the dimensioned drawings are only about 80% of the correct size. See Figure 1.

Chicago diag 

From the Chicago Guide. Car widths added for this article. If the image does not show on your browser, then its URL is: and the page is number 7. The truck at lower left has 4.9 foot body width; the moving car is 4.9 feet wide; the car at upper right is 4.9 feet wide; the car at right middle is 5.3 feet wide.

3.2.2 Door Zone Considerations

Chicago uses several different widths for parking lanes, which it marks with a stripe, and bike lanes, resulting in several different total widths. So does the AASHTO Guide. Where the parking lane is marked, AASHTO does not specify the lane width in its Guide for Bicycle Facilities (traditional parking lane width is 8'), but specifies a bike-lane width of 5' outside the marked parking lane. Where the parking lane is not marked, AASHTO specifies a minimum of 11' total where parking turnover is infrequent, up to 13' for areas with "substantial parking or turnover of parked cars is high (e.g. commercial areas)."

 Chicago uses 7' parking lane and 5' bike lane as its minimum standard, used only on 44' wide 2-lane streets and 55' wide 3-lane streets. Where additional width is present, Chicago splits the additional width between bike lane and parking lane until the total reaches13'. Assuming that the 44', 2-lane streets are residential streets, Chicago's standards are quite similar to AASHTO's.

 The question remains as to the reasonableness of these standards for avoiding opened car doors. With typical car body-widths of 5.8' to 6.1', typical door widths of 3.2', curb gap of 0.7', and bicycle handlebar width of 2', the minimum total lane width required to clear typical opened doors is 12'. For larger cars, it runs to 12.5' or more. For commercial vehicles it is about 14'.

The fact that some of the parked cars are narrower than others does not alter the width required, because the cyclist has to ride straight outside the danger zone of the widest vehicles present. If there are only a very few wide vehicles parked, as commercial vehicles in a residential zone, then the cyclist may steer to the left to clear those. For parking consisting largely of passenger cars only, 12' is the absolute minimum for generally safe clearance of doors, while 13' would avoid all but the obvious wide vehicle.

However, there is the issue of delineating the parking lane. If it is assumed that the bodies of the parked vehicles extend out to the delineated stripe, then the 5' bike lane is insufficient to clear almost any open doors. To do that within a total width of 12', the parking lane must be delineated as only 6.5' in width. That will handily accommodate passenger cars (drivers of large cars will have to park close to the curb), but will not accommodate commercial vehicles. However, commercial vehicles are obvious to the cyclist, who would have to steer a bit left to clear them safely in any case, regardless of the delineation of the parking lane.

 One favorable characteristic of buses needs comment. That is, that while buses are of maximum legal width, 8.3' (100"), they have no doors on the traffic side. In other words, a bus, wide as it is, does not project as far into the roadway as does the door zone of the typical automobile.


I would judge, on the basis of the design standards, that both AASHTO and Chicago hover about the absolute minimum while recognizing that more width is desirable for greater safety from opened doors. The complicated issue remains of how to define legal parking lateral position, as distance from the curb, or within a delineated parking lane, and the widths of parking lanes when delineated, and how to provide for the required exceptions when the parking lane is too narrow for some legal vehicles.

3.3 Logical Basis?

The logic of the whole is just the same as ever before. That is, bike lanes are striped where there is adequate width for cyclists to operate alongside motor traffic, but where is inadequate width to do this, there is no bike lane. On bikelaned streets where there is a demand for additional width for turn lanes or bus stops, the bike lane just stops, with some form of tapered dotted stripe indicating that the cyclist has to merge with motor traffic.

4 Conclusions About the Utility of Bikeways

We have learned four things from these studies.
  1. 1: The cyclist who rides at the left-hand edge of American-standard bike lanes is barely avoiding the open-door-zone of most vehicles, but will have to leave the bike lane to avoid the door zones of larger vehicles.
  2. 2: Europeans tend to exclude cyclists from roadways to a much greater extent than do Americans.
  3. 3: The Chicago Bike Lane Design Guide, though reasonably well done of its kind (although with pictorial scaling errors that make the designs appear much better than they are), demonstrates yet again that there is no reason to have bike lanes.
  4. King's paper, Bicycle Facility Selection: A Comparison of Approaches, aside from the data that he has collected from previously published sources, contains nothing more than the typical bike-planners' propaganda, based on common superstition all wrapped up in the appearance of valid engineering knowledge.

King's paper is based on the superstition that the only factor to be considered when determining the method by which cyclists are to be accommodated on a highway is the speed and volume of same-direction motor traffic. That is not only King's superstition, but the superstition on which all of his sources are based. (At least, I know of some examples and I have never heard of any exceptions.) The Chicago Guide demonstrates the truth of this conclusion, because it demonstrates, through its considerable detail, that bike lanes do nothing to alleviate the predominant traffic dangers to cyclists of crossing and turning movements by themselves or by motor traffic.

While Mr. King prates that he assumed that the design cyclists were "non-vehicular (casual, average & child) cyclists", there is nothing in his analysis to suggest that the facilities that he surveyed were safe or appropriate for such cyclists. He simply assumed that facilities such as he surveyed are safe and appropriate for incompetent cyclists, while other facilities are dangerous for such. Contrasted with this is the obvious conclusion to be drawn from Chicago's detailed designs, particularly concerning what those designs do not do. To use Chicago's bike lanes safely requires the full range of vehicular-cycling skills.

In short, the comparison of these two documents is another demonstration of what has been obvious for decades. That is, while bike-planning propaganda reads well because it reflects common superstition without considering the question of whether or not it has a scientific basis, when the details are examined, as in Chicago's detailed standard designs, bike-planning theory is shown to be empty of all scientific bases.

Return to: John Forester's Home Page                                            Up: Facilities